Notes on Macroeconomic Theory

Steve Williamson Dept. of Economics Washington University in St. Louis St. Louis, MO 63130 September 2006

Chapter 1 Simple Representative Agent Models
This chapter deals with the simplest kind of macroeconomic model, which abstracts from all issues of heterogeneity and distribution among economic agents. Here, we study an economy consisting of a representative firm and a representative consumer. As we will show, this is equivalent, under some circumstances, to studying an economy with many identical firms and many identical consumers. Here, as in all the models we will study, economic agents optimize, i.e. they maximize some objective subject to the constraints they face. The preferences of consumers, the technology available to firms, and the endowments of resources available to consumers and firms, combined with optimizing behavior and some notion of equilibrium, allow us to use the model to make predictions. Here, the equilibrium concept we will use is competitive equilibrium, i.e. all economic agents are assumed to be price-takers.

1.1
1.1.1

A Static Model
Preferences, endowments, and technology

There is one period and N consumers, who each have preferences given by the utility function u(c, ), where c is consumption and is leisure. Here, u(·, ·) is strictly increasing in each argument, strictly concave, and 1

2

CHAPTER 1. SIMPLE REPRESENTATIVE AGENT MODELS

twice differentiable. Also, assume that limc→0 u1 (c, ) = ∞, > 0, and lim →0 u2 (c, ) = ∞, c > 0. Here, ui (c, ) is the partial derivative with respect to argument i of u(c, ). Each consumer is endowed with one unit of time, which can be allocated between work and leisure. Each consumer also owns k0 units of capital, which can be rented to firms. N There are M firms, which each have a technology for producing consumption goods according to y = zf (k, n), where y is output, k is the capital input, n is the labor input, and z is a parameter representing total factor productivity. Here, the function f (·, ·) is strictly increasing in both arguments, strictly quasiconcave, twice differentiable, and homogeneous of degree one. That is, production is constant returns to scale, so that λy = zf (λk, λn), (1.1)

for λ > 0. Also, assume that limk→0 f1 (k, n) = ∞, limk→∞ f1 (k, n) = 0, limn→0 f2 (k, n) = ∞, and limn→∞ f2 (k, n) = 0.

1.1.2

Optimization

In a competitive equilibrium, we can at most determine all relative prices, so the price of one good can arbitrarily be set to 1 with no loss of generality. We call this good the numeraire. We will follow convention here by treating the consumption good as the numeraire. There are markets in three objects, consumption, leisure, and the rental services of capital. The price of leisure in units of consumption is w, and the rental rate on capital (again, in units of consumption) is r. Consumer’s Problem Each consumer treats w as being fixed, and maximizes utility subject to his/her constraints. That is, each solves max u(c, )
c, ,ks

A STATIC MODEL subject to c ≤ w(1 − ) + rks k0 0 ≤ ks ≤ N 0≤ ≤1 c≥0 3 (1.2) (1.3) states that the quantity of capital rented must be positive and cannot exceed what the consumer is endowed with.3) (1. Our restrictions on the utility function assure that there is a unique optimum which is characterized by the following first-order conditions. ui is the partial derivative of u(·. Now. ∂L = u1 − μ = 0 ∂c ∂L = u2 − μw = 0 ∂ ∂L k0 =w+r −w −c=0 ∂μ N Here. and (1.2) will hold with equality. so we can safely ignore this case. The optimization problem for the consumer is therefore much simplified. ) − u2 (w + r − w .6) N N . The above first-order conditions can be used to solve out for μ and c to obtain k0 k0 wu1 (w + r − w . (1. ) + μ(w + r − w − c). ·) with respect to argument i.4) is a similar condition for leisure. (1. (1. given that utility is increasing in consumption (more is preferred to less). we must have ks = k0 . ) = 0. and we can write down the following Lagrangian for the problem.5) is a nonnegativity constraint on consumption.2) is the budget constraint. and in equilibrium we will never have = 1. as then nothing would be produced.1. and (1. N Our restrictions on the utility function assure that the nonnegativity constraints on consumption and leisure will not be binding. (1. k0 L = u(c. ks is the quantity of capital that the consumer rents to firms.5) Here.1.4) (1. N where μ is a Lagrange multiplier.

c E (rk )/N 0 B D 1 leisure. l Figure 1.4 CHAPTER 1. in Figure 1.1: which solves for the desired quantity of leisure. n) − rk − wn]. the marginal rate of substitution of leisure for consumption equals the wage rate. a firm solves max[zf (k. . Equation (1. is tangent to the highest indifference curve. and k0 . the consumer’s budget constraint is ABD. That is. where the budget constraint. u1 i.1: Consumer's Optimization Problem A Consumption. u1 Firm’s Problem Each firm chooses inputs of labor and capital to maximize profits. in terms of w. Diagrammatically. k.1. SIMPLE REPRESENTATIVE AGENT MODELS Figure 1. r. which has slope −w.e. where an indifference curve has slope − u2 .6) can be rewritten as N u2 = w.n . and he/she maximizes utility at E. treating w and r as being fixed.

for example). A STATIC MODEL 5 and the first-order conditions for an optimum are the marginal product conditions zf1 = r. c. Secondly.3 Competitive Equilibrium A competitive equilibrium is a set of quantities.8).1. The representative firm chooses n and k optimally given w and r. the optimal scale of operation of the firm is indeterminate. That is. (1.1.7) zf2 = w. The first is that we do not need to be concerned with how the firm’s profits are distributed (through shares owned by consumers. due to the constant returns to scale assumption. Now. 1. (1.1. then we must have zf (k. representative firm). which satisfy the following properties. since (1.8) where fi denotes the partial derivative of f (·. .9) Equations (1. n) − rk − wn = 0 (1. 2. (1.1) with respect to λ. differentiating (1. and setting λ = 1. and (1. k. Euler’s law holds. n) = zf1 k + zf2 n. given that the function f (·.9) then imply that maximized profits equal zero. It therefore makes no difference for our analysis to simply consider the case M = 1 (a single. This has two important consequences.10) also holds for k = λk ∗ and n = λn∗ for any λ > 0.7). n. But. ·) is homogeneous of degree one. Markets clear. suppose k∗ and n∗ are optimal choices for the factor inputs. 3. . we get zf (k. ·) with respect to argument i. 1. Each consumer chooses c and optimally given w and r.10) for k = k∗ and n = n∗ . (1. as the number of firms will be irrelevant for determining the competitive equilibrium. and prices w and r.

then the additional market is also in equilibrium. supply equals demand in each market given prices. Walras’ law states that the value of excess demand across markets is always zero. if any 2 of (1.12) (1. we have Nc − y + w[n − N(1 − )] + r(k − k0 ) = 0.7). and were distributed lump-sum to consumers. Now. then (1.13). We can therefore drop one market-clearing condition in determining competitive equilibrium prices and quantities. and (1. but from the consumer’s budget constraint. given (3). N(1 − ) = n y = Nc k0 = k (1. (1. Here. in this context our results would not be any different if there were many firms .6 CHAPTER 1.11).8). n. In a competitive equilibrium. SIMPLE REPRESENTATIVE AGENT MODELS Here.13) hold. and this then implies that.6). the following conditions then hold. (1.11) (1. and (1.11). But now. It should be apparent here that the number of consumers. Competitive equilibrium might seem inappropriate when there is one consumer and one firm. and r.12). the market for consumption goods. (1. and simply analyze an economy with a single representative consumer. These are five equations in the five unknowns . if there are M markets and M − 1 of those markets are in equilibrium. the total value of excess demand across markets is Nc − y + w[n − N(1 − )] + r(k − k0 ).14) Note that (1. and we can solve for c using the consumer’s budget constraint. N.13) That is. and the fact that profit maximization implies zero profits. (1. we eliminate (1. is virtually irrelevant to the equilibrium solution. and the market for rental services of capital. The competitive equilibrium is then the solution to (1.12). there are three markets: the labor market. w. Equation (1.14) implies that the third market-clearing condition holds. but as we have shown.14) is simply Walras’ law for this model.14) would hold even if profits were not zero. k. (1. so for convenience we can set N = 1.

zf2 (k0 . 1 − ) (1. A STATIC MODEL 7 and many consumers.15) Given the solution for . max u(c.1. all in a way that makes the consumer as well off as possible.1. Here.16) (1. We can substitute in equation (1. 1 − ). w. ) − u2 (zf (k0 . 1 − ) (1.19) . k. ) c. and c. ) = 0 (1. 1 − ) = w n=1− k = k0 c = zf (k0 . can force the consumer to supply the appropriate quantity of labor. generally. subject to c = zf (k0 . and then distributes consumption goods to the consumer. 1. The social planner determines a Pareto optimum by solving the following problem. It helps to think in terms of a fictitious social planner who can dictate inputs to production by the representative firm.1. is defined to be some allocation (an allocation being a production plan and a distribution of goods across economic agents) such that there is no other allocation which some agents strictly prefer which does not make any agents worse off. 1 − )u1 (zf (k0 . 1 − ). 1 − ) = r zf2 (k0 .4 Pareto Optimality A Pareto optimum. we do not have to worry about the allocation of goods across agents.6) to obtain an equation which solves for equilibrium .17) It is not immediately apparent that the competitive equilibrium exists and is unique. but we will show this later. since we have a single agent.18) (1. n. zf1 (k0 . we then substitute in the following equations to obtain solutions for r.

and differentiate with respect to to obtain the following first-order condition for an optimum. In Figure 1.18). income taxes and sales taxes). and the right side is the marginal rate of substitution of consumption for leisure.8 CHAPTER 1. zf2 (k0 . u1 where the left side of the equation is the marginal rate of transformation. if we can . 1 − ). and the solution we get for c from the social planner’s problem by substituting in the constraint will yield the same solution as from (1. In macroeconomics. is equal to − u2 . u1 In more general settings. In a competitive equilibrium.2. SIMPLE REPRESENTATIVE AGENT MODELS Given the restrictions on the utility function. Note that we can rewrite (1. Any Pareto optimum can be supported as a competitive equilibrium with an appropriate choice of endowments. ·) is strictly concave and f (·. (Second Welfare Theorem). AB is equation (1. since u(·. 1 − ). and the general idea goes back as far as Adam Smith’s Wealth of Nations. A competitive equilibrium is Pareto optimal (First Welfare Theorem).20) are identical. the competitive equilibrium and the Pareto optimum are identical here. Further.g. ·) is strictly quasiconcave. 1. where the highest indifference curve is tangent to the production possibilities frontier.20) Note that (1. ] = 0 (1. there is a unique Pareto optimum. That is.20) as zf2 = u2 . and absence of distorting taxes (e. The First Welfare Theorem is quite powerful.19) and the Pareto optimum is at D. we can simply substitute using the constraint in the objective function. ] − u2 [zf (k0 . The non-technical assumptions required for (1) and (2) to go through include the absence of externalities. 2. it is true under some restrictions that the following hold. 1 − )u1 [zf (k0 . the representative consumer faces budget constraint EFB and maximizes at point D where the slope of the budget line. and the competitive equilibrium is also unique. completeness of markets.15) and (1. −w.

and then solving for prices.16) and (1.1. and then finding w and r from the appropriate marginal conditions. A STATIC MODEL 9 Figure 1. l Figure 1.17). in the above example. c F B 1 leisure. . a competitive equilibrium could be obtained by first solving for c and from the social planner’s problem. business cycles) using a competitive equilibrium model in which the First Welfare Theorem holds. rather than solving simultaneously for prices and quantities using marketclearing conditions. the equivalence of competitive equilibria and Pareto optima in representative agent models is useful for computational purposes.2: successfully explain particular phenomena (e. (1. Using this approach does not make much difference here. but in computing numerical solutions in dynamic models it can make a huge difference in the computational burden.1. For example. In addition to policy implications. we can then argue that the existence of such phenomena is not grounds for government intervention.2: Pareto Optimum and Competitive Equilibrium E A D Consumption. That is.g. it can be much easier to obtain competitive equilibria by first solving the social planner’s problem to obtain competitive equilibrium quantities.

Note that c1−γ − 1 lim = lim γ→1 1 − γ γ→1 d [e(1−γ) log c − dγ d (1 − γ) dγ 1] = log c.1. the production function is Cobb-Douglas. For the utility function. we have α w = [(1 − α)1−α (zk0 )] α+(1−α)γ γ (1.23) From (1. where 0 < α < 1. ) = + .5 Example Consider the following specific functional forms. SIMPLE REPRESENTATIVE AGENT MODELS 1.22) and (1. That is. For the production technology. However. The social planner’s problem here is then α [zk0 (1 − )1−α ]1−γ − 1 + max 1−γ ( ) . using L’Hospital’s Rule.10 CHAPTER 1. Solving for c.17). 1 (1.23) clearly c and w are increasing in z and k0 .21) the effects . increases in productivity and in the capital stock increase aggregate consumption and real wages. we use c1−γ − 1 u(c. from (1. this is also the competitive equilibrium solution. 1−γ where γ > 0 measures the degree of curvature in the utility function with respect to consumption (this is a “constant relative risk aversion” utility function).22) and from (1.21) As in the general case above. use f (k. That is. and the solution to this problem is α = 1 − [(1 − α)(zk0 )1−γ ] α+(1−α)γ 1 (1.19). we get α c = [(1 − α)1−α (zk0 )] α+(1−α)γ . n) = kα n1−α . from equation (1.

and the first-order condition for a maximum is −zu1 [z(1 − ).6 Linear Technology . 1. Which way the effect goes depends on whether γ < 1 or γ > 1. corresponds to an increase in the real wage faced by the consumer. this model can deliver the result that employment and output move in opposite directions. but the effects are just the opposite if γ > 1. In the data.1. (1. where fluctuations are driven by technology shocks (changes in z). ]. and aggregate employment are mutually positively correlated. ∂n so that an increase in productivity.1. in a somewhat more general sense than the above example.e. we cannot solve explicitly for . ] = 0. and shows. z. To determine the effect of an increase . but still consider a constant returns to scale technology with labor being the only input. With γ < 1. we eliminate capital. why a productivity shock might give a decrease or an increase in employment. If we want to treat this as a simple model of the business cycle. The social planner’s problem for this economy is then max u[z(1 − ). and an increase in employment. as is the case in the data. an increase in z or in k0 will result in a decrease in leisure. To make things clearer.Comparative Statics This section illustrates the use of comparative statics. i. these results are troubling. that the real wage will be procyclical (it goes up when output goes up). in contrast to the example. A STATIC MODEL 11 on the quantity of leisure (and therefore on employment) are ambiguous. aggregate output. we consider a simplified technology. aggregate consumption. Note however. but note that the equilibrium real wage is w= ∂y = z. ] + u2 [z(1 − ).24) Here.1. y = zn. However.

d u1 < 0.12 CHAPTER 1. = dz z 2 u11 − 2zu12 + u22 (1. and −zu1 (c. the social optimum (also the competitive equilibrium) is at E initially.27) Totally differentiating (1. it d d is easy to construct examples where dz > 0. ) = h. The ambiguity here arises from opposing income and substitution effects. In fact. u(c.) dz is just the remainder. z2 > z1 . with no change in but higher c. AB denotes the resource constraint faced by the social planner. and BD is the resource constraint with a higher level of productivity.25) then. concavity of the utility function implies that the denominator in (1. apply the implicit function theorem and totally differentiate (1. Algebraically. dz z 2 u11 − 2zu12 + u22 .26) where h is a constant. (subst.27) with respect to c and . and where dz < 0.25) We then have Now.) = 2 dz z u11 − 2zu12 + u22 From (1.) as follows. As shown. SIMPLE REPRESENTATIVE AGENT MODELS in z on . and usd ing (1. d u1 + z(1 − )u11 − u21 (1 − ) . the income effect d (inc.3. In Figure 1.25) is negative.24) to get [−u1 − z(1 − )u11 + u21 (1 − )]dz +(z 2 u11 − 2zu12 + u22 )d = 0. we can solve for the substitution effect dz (subst. and his/her response can be decomposed into a substitution effect (E to G) and an income effect (G to F). but we cannot sign the numerator. and at F after the increase in productivity.e. the representative consumer faces a higher real wage. i. Effectively. (1. ) + u2 (c. c = z1 (1 − ).27) to simplify.) = > 0. d z(1 − )u11 − u21 (1 − ) (inc. ) = 0 (1. we can determine the substitution effect on leisure by changing prices and compensating the consumer to hold utility constant.26) and (1.

as it is in the data. we require a substitution effect that is large relative to the income effect. as we will see later. c z1 A G F E B 1 leisure. l Figure 1. and finances these purchases through lump-sum taxes on the representative consumer. 1. and let τ be total taxes.2. preferences and substitution effects are very important in equilibrium theories of the business cycle.3: Effect of a Productivity Shock z2 D Consumption. which is treated as being exogenous. Therefore. That is. which increases the real wage and output. must result in a decrease in leisure in order for employment to be procyclical. The government . GOVERNMENT 13 Figure 1. we introduce a government sector into our simple static model in the following manner. The government makes purchases of consumption goods. in order for a model like this one to be consistent with observation. a productivity shock.1. Let g be the quantity of government purchases. In general.3: provided is a normal good.2 Government So that we can analyze some simple fiscal policy issues.

. . n Therefore. i. g) = u(c. ) + v(g) Given that utility is separable in this fashion. unless we wish to consider the optimal determination of government purchases. ) c. but it simplifies matters. w(c. A competitive equilibrium consists of quantities. subject to and the first-order condition for an optimum is −wu1 + u2 = 0.e.e. we can assume v(g) = 0. c = w(1 − ) − τ. g = τ. The representative consumer chooses c and to maximize utility. given w and τ. and does not make much difference for the analysis. This is clearly unrealistic (in most cases). For example. The representative firm’s profit maximization problem is max(z − w)n. SIMPLE REPRESENTATIVE AGENT MODELS budget must balance.28) We assume here that the government destroys the goods it purchases. Given this.14 CHAPTER 1. n. we could allow government spending to enter the consumer’s utility function in the following way. i. the firm’s demand for labor is infinitely elastic at w = z. which satisfy the following conditions: 1. c. w. Here. assume a technology of the form y = zn. and a price. . As in the previous section. this would make no difference for the analysis. (1. labor is the only factor of production. and g is exogenous. the consumer’s optimization problem is max u(c. and τ.

In Figure 1. ) c. subject to c + g = z(1 − ) Substituting for c in the objective function.2.5. Given the government budget constraint.1. The competitive equilibrium and the Pareto optimum are equivalent here. and maximizing with respect to . there is an increase in taxes. shifting in the resource constraint. 4. the first-order condition for this problem yields an equation which solves for : −zu1 [z(1 − ) − g. Algebraically.4. there is crowding out of private consumption. and the Pareto optimum (competitive equilibrium) is D. Note that the slope of the resource constraint is −z = −w.29) and the equation c = z(1 − ) − g and solve to obtain d −zu11 + u12 = 2 <0 dg z u11 − 2zu12 + u22 dc zu12 − u22 = 2 <0 dg z u11 − 2zu12 + u22 (1. (1.28).30) . totally differentiate (1.29) In Figure 1. The social planner’s problem is max u(c. GOVERNMENT 15 2. is satisfied. ] + u2 [z(1 − ) − g. We can now ask what the effect of a change in government expenditures would be on consumption and employment. which represents a pure income effect for the consumer. given w. but note that the decrease in consumption is smaller than the increase in government purchases. ] = 0. (1. 3. The representative firm chooses n to maximize profits. the economy’s resource constraint is AB. Markets for consumption goods and labor clear. so that output increases. Thus. quantities of both goods will decrease. g increases from g1 to g2 . Given that leisure and consumption are normal goods. as in the version of the model without government. The government budget constraint.

The dynamics are restricted to the government’s financing decisions.e.30) also implies that dy < 1. SIMPLE REPRESENTATIVE AGENT MODELS Figure 1. normal goods. the social planner’s problem will break down into a series of static optimization problems. Note that (1. i. ∞ X t=0 β t u(ct .3 A “Dynamic” Economy We will introduce some simple dynamics to our model in this section. the “balanced budget dg multiplier” is less than 1. there are really no dynamic elements in terms of real resource allocation.0) 1 leisure.e. we deal with an infinite horizon economy. Here. l B -g Figure 1. t ). This model will be useful for studying the effects of changes in the timing of taxes.4: Linear Production and Government Spending z-g A Consumption. respectively.16 CHAPTER 1.e. if leisure and consumption are. .4: Here. the inequalities hold provided that −zu11 + u12 > 0 and zu12 − u22 > 0. where the representative consumer maximizes time-separable utility. c D (0. 1. i. i.

0 < β < 1.5: where β is the discount factor. 2. . Government purchases are financed through lump-sum taxation and by issuing one-period government bonds. 1. 2. Each period. the consumer is endowed with one unit of time... The government purchases gt units of consumption goods in period t. we have β = 1+δ . where rt+1 is the one-period interest rate.0) -g 1 1 leisure.1.3. b0 = 0.31) states that government purchases plus principal and interest on the government debt is equal to tax revenues plus new bond issues. (1. Letting δ denote the dis1 count rate...5: Increase in Government Spending z-g 1 z-g 2 Consumption.31) t = 0. t = 0. where bt is the number of one-period bonds issued by the government in period t − 1. . A “DYNAMIC” ECONOMY 17 Figure 1. 1. and these purchases are destroyed. c A B (0. l -g 2 Figure 1. There is a representative firm which produces output according to the production function yt = zt nt ... Here. The government budget constraint is gt + (1 + rt )bt = τt + bt+1 . . A bond issued in period t is a claim to 1+rt+1 units of consumption in period t+1. Equation (1. where δ > 0.

.ct . . 2. must equal zero in the limit. i=1 (1 + ri ) Qt ct = w0 (1 − i=1 (1 + ri ) 0) − τ0 + ∞ X wt (1 − t ) − τt .. 2. i=1 (1 + ri ) u1 (c0 . 0 ) − λw0 = 0 λwt = 0.33) n→∞ i=1 (1 + ri ) ct = wt (1 − t ) − τt − st+1 + (1 + rt )st .. SIMPLE REPRESENTATIVE AGENT MODELS The optimization problem solved by the representative consumer is {st+1 . s0 = 0.32) gives c0 + ∞ X t=1 Now. discounted to t = 0. 1. maximizing utility subject to the above intertemporal budget constraint.32) which states that the quantity of debt. we obtain the following first-order conditions. t ) (1. (1. which come due in period t + 1...34) Qt i=1 (1 + ri ) t=1 Here. We then obtain u2 (ct . t }t=0 .. t ) − Qt β t u2 (ct . We will assume that sn lim Qn−1 = 0. where st+1 is the quantity of bonds purchased by the consumer in period t. t ) subject to t = 0.18 CHAPTER 1. 2. 3.” and implies that we can write the sequence of budget constraints.35) u2 (c0 . t = 1. t ) = wt . u1 (ct . Here. (1. 0 ) − λ = 0 . Repeated substitution using (1.32) as a single intertemporal budget constraint. 3. β t u1 (ct . we permit the representative consumer to issue private bonds which are perfect substitutes for government bonds. This condition rules out infinite borrowing or “Ponzi schemes. . max ∞ ∞ X t=0 β t u(ct . (1. .. (1. t = 1. λ is the Lagrange multiplier associated with the consumer’s intertemporal budget constraint. t ) − Qt λ = 0.

4. Qt i=1 (1 + ri ) i=1 (1 + ri ) t=1 (1. τt }∞ . A competitive equilibrium consists of quantities. The representative firm simply maximizes profits in each period.e. n t and labor demand. Given {gt }∞ . and βu1 (ct+1 . the marginal rate of substitution of leisure for consumption in any period equals the wage rate.1.31). Markets for consumption goods. {ct . the present discounted value of government purchases equals the present discounted value of tax revenues. 2. nt ... t=0 t=0 3. . t ) 1 + rt+1 (1.37) = nt . t . rt+1 }∞ . bt+1 . i.38) . t=0 and prices {wt . and 1− t (1.. }∞ optimally given {τt } and {wt . t+1 ) 1 = .36) i. Firms choose {nt }∞ optimally given {wt }∞ . τt }∞ satisfies the sequence of government t=0 t=0 budget constraints (1. 1. Walras’ law permits us to drop the consumption goods market from consideration. is perfectly elastic at wt = zt . rt+1 }∞ satisfying the following conditions. giving us two market-clearing conditions: st+1 = bt+1 . Now. 2.3.33) and (1. t=0 t=0 2. st+1 . {bt+1 . Consumers choose {ct .e.37) imply that we can write the sequence of government budget constraints as a single intertemporal government budget constraint (through repeated substitution): g0 + ∞ X t=1 i.e.. it solves max(zt − wt )nt . 1.e. u1 (ct . t = 0. t=0 1. (1. labor. t . st+1 . A “DYNAMIC” ECONOMY 19 i. . Now. the intertemporal marginal rate of substitution of consumption equals the inverse of one plus the interest rate. nt .. and bonds clear. t = 0. since the government Qt ∞ X gt τt = τ0 + .

suppose that {wt . (1.39) Qt i=1 (1 + ri ) t=1 max ∞ X t=0 β t u[zt (1 − t ) − gt . (1. { t }∞ t=0 Qt ct = w0 (1 − i=1 (1 + ri ) 0 ) − g0 + ∞ X wt (1 − t ) − gt . Another way to show the Ricardian equivalence result here comes from computing the competitive equilibrium as the solution to a social planner’s problem.e.. . leisure. the t=0 representative firm’s choices are invariant. t = 0. Then. and where the dynamics are more complicated. and we can solve for ct from ct = zt (1 − t ). and the first-order conditions for an optimum are −zt u1 [zt (1 − t ) − gt . . (1. Here.39) implies that the optimizing choices given those prices remain optimal given any sequence {τt }∞ satisfying (1. i.36) determine prices.40) solves for t . and private saving increases by an equal amount. For example. The government issues more debt today to finance an increase in the government deficit. since the representative consumer saves more to pay the higher taxes in the future. 1. Also.35) and (1.. we can use (1.. he/she knows that the government will have to make this up with higher future taxes. holding the path of government purchases constant. it is clear that the timing of taxes is irrelevant to determining the competitive equilibrium. 2. . Here.20 CHAPTER 1.. Some assumptions which are critical to the Ricardian equivalence result are: . (1. 2..40) t = 0. all that is relevant for the determination of consumption. This is a version of the Ricardian Equivalence Theorem. (1. t ] This breaks down into a series of static problems. t=0 Then. That is. and the timing of taxes is irrelevant. and prices.34) to obtain c0 + ∞ X t=1 Now. rt+1 }∞ are competitive equilibrium prices. is the present discounted value of government purchases. SIMPLE REPRESENTATIVE AGENT MODELS budget constraint must hold in equilibrium. 1.38) to substitute in (1.38). t ] = 0. if the representative consumer receives a tax cut today. t ] + u2 [zt (1 − t ) − gt . though Ricardian equivalence holds in much more general settings where competitive equilibria are not Pareto optimal.

Consumers are infinite-lived. A “DYNAMIC” ECONOMY 1. Capital markets are perfect.e. Taxes are lump sum 2. . the present discounted value of each individual’s tax burden is unaffected by changes in the timing of aggregate taxation. i. 21 3. the interest rate at which private agents can borrow and lend is the same as the interest rate at which the government borrows and lends.3.1. There are no distributional effects of taxation. That is. 4.

22 CHAPTER 1. SIMPLE REPRESENTATIVE AGENT MODELS .

government debt policy can be used as a vehicle for redistributing wealth among generations and inducing optimal savings behavior. 23 .Chapter 2 Growth With Overlapping Generations This chapter will serve as an introduction to neoclassical growth theory and to the overlapping generations model. competitive equilibria need not be Pareto optimal. There was a resurgence in interest in the overlapping generations model as a monetary paradigm in the late seventies and early eighties. As a result. and agents currently alive cannot trade with the unborn. building on the overlapping generations construct introduced by Samuelson (1956). resources may not be allocated optimally among generations. particularly at the University of Minnesota (see for example Kareken and Wallace 1980). The particular model introduced in this chapter was developed by Diamond (1965). and Ricardian equivalence does not hold. the timing of taxes and the size of the government debt matters. A key feature of the overlapping generations model is that markets are incomplete. in a sense. in that economic agents are finite-lived. Thus. Samuelson’s paper was a semi-serious (meaning that Samuelson did not take it too seriously) attempt to model money. Without government intervention. However. but it has also proved to be a useful vehicle for studying public finance issues such as government debt policy and the effects of social security systems. and capital accumulation may be suboptimal.

These last two conditions on the marginal rate of substitution will imply that each consumer will always wish to consume positive amounts when young and when old. 1. and renting capital and hiring labor as inputs. Consumption goods can be converted one-for-one into capital. and defining v(cy . Each period.24CHAPTER 2.. . and there is no depreciation. and zero units in the second period. co ) ≡ ∂u ∂cy ∂u ∂co . co ). In period 0 there are some old consumers alive who live for one period and are collectively endowed with K0 units of capital. t t+1 where cy denotes the consumption of a young consumer in period t and t co is the consumption of an old consumer. The representative firm maximizes profits by producing consumption goods.1 The Model This is an infinite horizon model where time is indexed by t = 0. The technology is given by Yt = F (Kt .. ·) is strictly t increasing in both arguments. strictly concave. GROWTH WITH OVERLAPPING GENERATIONS 2. Preferences for a consumer born in period t. 2. ∞. and each is endowed with one unit of labor in the first period of life. . The population evolves according to Lt = L0 (1 + n)t . Lt two-period-lived consumers are born. assume that limcy →o v(cy ... The investment technology works as follows. The initial old seek to maximize consumption in period 0.. 2. Young agents sell their labor to firms and save in the form of capital accumulation. are given by u(cy .. t = 0.1) where L0 is given and n > 0 is the population growth rate. co ) = ∞ for co > 0 and limco →o v(cy . Assume that u(·. Lt ). and old agents rent capital to firms and then convert the capital into consumption goods which they consume. (2. co ) = 0 for cy > 0. . and vice-versa. Capital constructed in period t does not become productive until period t + 1. 1.

We can then use (2.3) Definition 1 A Pareto optimal allocation is a sequence {cy .3) and the property that there exists no other allocation {ˆy . i. We will confine attention to allocations where all young agents in a given period are treated identically. capital accumulation.2) is the quantity of goods available in period t.. 2.2. and all old agents in a given period receive the same consumption. co ) ≥ u(cy .2) where the left hand side of (2. co . consumption goods produced plus the capital that is left after production takes place. ·) is strictly increasing.1) to rewrite (2. respectively. 1).2. The right hand side is the capital which will become productive in period t + 1 plus the consumption of the young.3) and ct ˆt ˆ t=0 u(ˆy .2 Optimal Allocations As a benchmark. co . twice differentiable. 3. The resource constraint faced by the social planner in period t is F (Kt . with strict inequality in at least one instance. . and homogeneous of degree one. we will first consider the allocations that can be achieved by a social planner who has control over production. 1.e.2) as f (kt ) + kt = (1 + n)kt+1 + cy + t co t 1+n (2. Define kt ≡ Ktt (the capital/labor ratio or per-capita capital stock) and L f (kt ) ≡ F (kt .. kt+1 }∞ which satisfies (2. Assume that the production function F (·. Lt ) + Kt = Kt+1 + cy Lt + co Lt−1 .. co ) ct ˆt+1 t t+1 co ≥ co ˆ1 1 for all t = 0. t t (2. plus consumption of the old. Thus. In the long run. and the distribution of consumption goods between the young and the old. kt+1 }∞ t t t=0 satisfying (2. . 2. this model will have the property that per-capita quantities converge to constants. OPTIMAL ALLOCATIONS 25 where Yt is output and Kt and Lt are the capital and labor inputs. strictly quasi-concave. it proves to be convenient to express everything here in per-capita terms using lower case letters.

Second. a Pareto optimal allocation is a feasible allocation such that there is no other feasible allocation for which all consumers are at least as well off and some consumer is better off. it is somewhat complicated (for our purposes) to derive Pareto optimal allocations here. and k to solve max u(cy . cy . and co are t t constants. While Pareto optimality is the appropriate notion of social optimality for this model. co . We will take a shortcut by focusing attention on steady states. but the two allocations may be Pareto non-comparable along the path to the steady state. and co = co . That is.6) or (marginal product of capital equal to n).4) 1+n Substituting for co in the objective function using (2. where k. (2. the planner finds the . GROWTH WITH OVERLAPPING GENERATIONS That is. First. one steady state may dominate another in terms of the welfare of consumers once the steady state is achieved.26CHAPTER 2. the planner chooses cy . (2.k subject to The first-order conditions for an optimum are then u1 − (1 + n)u2 = 0.5) u2 (intertemporal marginal rate of substitution equal to 1 + n) and f 0 (k) = n (2. there may not be a feasible path which leads from k0 to a particular steady state. co ) co . [1 + n][f (k) − nk − cy ]) y f (k) − nk = cy + c . First. where kt = k.2). given the feasibility condition. Note that the planner’s problem splits into two separate components.4). u1 =1+n (2. We need to be aware of two potential problems here. The problem for the social planner is to maximize the utility of each consumer in the steady state. cy = cy . we then solve the following max u(cy .

i. 2. from (2.3. and then allocates consumption between the young and the old according to (2. forecasts cannot be incorrect in equilibrium if agents have rational expectations. wt and rt+1 . co ) t t+1 subject to cy = wt − st t co = st (1 + rt+1 ) t+1 (2. k is chosen to maximize the size of the budget set for the consumer in the steady state. no one makes systematic forecasting errors.8) Here.7) (2.5). COMPETITIVE EQUILIBRIUM 27 capital-labor ratio which maximizes the steady state quantity of resources. wt is the wage rate.1 Young Consumer’s Problem A consumer born in period t solves the following problem. we can think of this as a rational expectations or perfect foresight equilibrium.2. 2. and to ask whether a competitive equilibrium achieves the steady state social optimum characterized in the previous section. At time t the consumer is assumed to know rt+1 .1. where each consumer forecasts future prices. Since there is no uncertainty here.co . Note that the capital rental rate plays the role of an interest rate here. as being fixed. and optimizes based on those forecasts. . and st is saving when young.e. and then consumption is allocated between the young and the old to achieve the tangency between the aggregate resource constraint and an indifference curve at point A. Equivalently. we wish to determine the properties of a competitive equilibrium.6).st t t+1 max u(cy .3. forecasts are correct.3 Competitive Equilibrium In this section. rt is the capital rental rate. In equilibrium. The consumer chooses savings and consumption when young and old treating prices. cy . In Figure 2.

cy .1: Optimal Steady State in the OG Model (1+n)(f(k)-nk) consumption when old. co A f(k)-nk consumption when young.Figure 2.

. (ii) firm optimization.10) Note that (2.e.e. st (1 + rt+1 ))(1 + rt+1 ) = 0 (2. however the sign of ∂rt+1 is indeterminate due to opposing income and substitution effects. (2. st . GROWTH WITH OVERLAPPING GENERATIONS Substituting for cy and co in the above objective function using t t+1 (2. we have ∂wt > 0. i.28CHAPTER 2. 1.3 Competitive Equilibrium Definition 2 A competitive equilibrium is a sequence of quantities. (2. ..2 Representative Firm’s Problem The firm solves a static profit maximization problem Kt . .3.. 2. we can determine optimal savings as a function of prices st = s(wt . Lt ) − wt = 0. F2 (Kt .9) which determines st . st }∞ and a sequence of prices {wt . given the initial capital-labor ratio k0 . st (1 + rt+1 )) + u2 (wt − st . Given that consumption when young and consumption when old ∂s ∂s are both normal goods. we can rewrite these marginal conditions as f 0 (kt ) − rt = 0.12) 2. The first-order conditions for a maximum are the usual marginal conditions F1 (Kt .9) can also be rewritten as u1 = 1 + rt+1 . rt+1 ). ·) is homogeneous of degree 1.7) and (2. (2. the first-order condition for an optimum is then −u1 (wt − st . rt }∞ .11) f (kt ) − kt f 0 (kt ) − wt = 0. Lt ) − rt = 0. (iii) market clearing. Lt ) − wt Lt − rt Kt ]. the inu2 tertemporal marginal rate of substitution equals one plus the interest rate.8) to obtain a maximization problem with one choice variable. in each period t = 0. 2.3. Since F (·.Lt max[F (Kt . which satisfy (i) cont=0 t=0 sumer optimization. {kt+1 . i.

and 0 < α < 1.15) Given the Cobb-Douglass production function. (2.11) and (2. and Walras’ law tells us that we can drop one marketclearing condition.12).11) and (2. which essentially determines the supply of capital. from (2. Here. capital rental.13) and we can substitute in (2. We can then solve for {st }∞ from (2. γ > 0.14) Here.12) to get kt+1 (1 + n) = s(f (kt ) − kf 0 (kt ). Consumer optimization is summarized by equation (2. we have f (k) = γk α and f 0 (k) = γαk α−1 . as period t savings is equal to the capital that will be rented in period t+1.13) for wt and rt+1 from (2. and consumption goods. Once we have the equilibrium sequence of capitalt=1 labor ratios.2. for labor. rt+1 ).4. we have three markets.10). the first-order conditions from the firm’s optimization problem give α−1 rt = γαkt . (2. s t and solving this problem we obtain the optimal savings function st = β wt . (2.10).12).11) and (2.16) . co ) = ln cy + β ln co .11) and (2. solves for {kt }∞ . (2. a young agent solves max[ln(wt − st ) + β ln[(1 + rt+1 )st )]. The demands for capital and labor are determined implicitly by equations (2. 1+β (2. and in turn for consumption allocations. L) = γK α L1−α . The supply of labor by consumers is inelastic. given k0 .4 An Example Let u(cy . The equilibrium condition for the capital rental market is then kt+1 (1 + n) = s(wt . and F (K. we can solve for prices from (2.12). It will be convenient here to drop the consumption goods market from consideration. where β > 0. Therefore. f 0 (kt+1 )).14) is a nonlinear first-order difference equation which. AN EXAMPLE 29 Here. t=0 2.

30CHAPTER 2. GROWTH WITH OVERLAPPING GENERATIONS
α wt = γ(1 − α)kt .

(2.17)

Then, using (2.14), (2.15), and (2.17), we get kt+1 (1 + n) = β α γ(1 − α)kt . (1 + β) (2.18)

Now, equation (2.18) determines a unique sequence {kt }∞ given k0 t=1 (see Figure 2m) which converges in the limit to k∗ , the unique steady state capital-labor ratio, which we can determine from (2.18) by setting kt+1 = kt = k∗ and solving to get βγ(1 − α) k = (1 + n)(1 + β)

"

#

1 1−α

.

(2.19)

Now, given the steady state capital-labor ratio from (2.19), we can solve for steady state prices from (2.16) and (2.17), that is r∗ = α(1 + n)(1 + β) , β(1 − α)
"

βγ(1 − α) w∗ = γ(1 − α) (1 + n)(1 + β)

#

α 1−α

.

We can then solve for steady state consumption allocations, cy = w∗ − co = β w∗ w∗ = , 1+β 1+β

β w∗ (1 + r∗ ). 1+β In the long run, this economy converges to a steady state where the capital-labor ratio, consumption allocations, the wage rate, and the rental rate on capital are constant. Since the capital-labor ratio is constant in the steady state and the labor input is growing at the rate n, the growth rate of the aggregate capital stock is also n in the steady state. In turn, aggregate output also grows at the rate n. ˆ Now, note that the socially optimal steady state capital stock, k, is determined by (2.6), that is ˆ γαk α−1 = n,

2.5. DISCUSSION or
µ ¶

31

αγ ˆ k= n

1 1−α

.

(2.20)

ˆ Note that, in general, from (2.19) and (2.20), k∗ 6= k, i.e. the competitive equilibrium steady state is in general not socially optimal, so this economy suffers from a dynamic inefficiency. There may be too little or too much capital in the steady state, depending on parameter values. ˆ That is, suppose β = 1 and n = .3. Then, if α < .103, k ∗ > k, and if ∗ ˆ α > .103, then k < k.

2.5

Discussion

The above example illustrates the dynamic inefficiency that can result in this economy in a competitive equilibrium.. There are essentially two problems here. The first is that there is either too little or too much capital in the steady state, so that the quantity of resources available to allocate between the young and the old is not optimal. Second, the steady state interest rate is not equal to n, i.e. consumers face the “wrong” interest rate and therefore misallocate consumption goods over time; there is either too much or too little saving in a competitive equilibrium. The root of the dynamic inefficiency is a form of market incompleteness, in that agents currently alive cannot trade with the unborn. To correct this inefficiency, it is necessary to have some mechanism which permits transfers between the old and the young.

2.6

Government Debt

One means to introduce intergenerational transfers into this economy is through government debt. Here, the government acts as a kind of financial intermediary which issues debt to young agents, transfers the proceeds to young agents, and then taxes the young of the next generation in order to pay the interest and principal on the debt. Let Bt+1 denote the quantity of one-period bonds issued by the government in period t. Each of these bonds is a promise to pay 1+rt+1

32CHAPTER 2. GROWTH WITH OVERLAPPING GENERATIONS units of consumption goods in period t + 1. Note that the interest rate on government bonds is the same as the rental rate on capital, as must be the case in equilibrium for agents to be willing to hold both capital and government bonds. We will assume that Bt+1 = bLt , (2.21)

where b is a constant. That is, the quantity of government debt is fixed in per-capita terms. The government’s budget constraint is Bt+1 + Tt = (1 + rt )Bt , (2.22)

i.e. the revenues from new bond issues and taxes in period t, Tt , equals the payments of interest and principal on government bonds issued in period t − 1. Taxes are levied lump-sum on young agents, and we will let τt denote the tax per young agent. We then have Tt = τt Lt . A young agent solves max u(wt − st − τt , (1 + rt+1 )st ), s
t

(2.23)

where st is savings, taking the form of acquisitions of capital and government bonds, which are perfect substitutes as assets. Optimal savings for a young agent is now given by st = s(wt − τt , rt+1 ). (2.24)

As before, profit maximization by the firm implies (2.11) and (2.12). A competitive equilibrium is defined as above, adding to the definition that there be a sequence of taxes {τt }∞ satisfying the government t=0 budget constraint. From (2.21), (2.22), and (2.23), we get τt =
µ

rt − n b 1+n

(2.25)

The asset market equilibrium condition is now kt+1 (1 + n) + b = s(wt − τt , rt+1 ), (2.26)

That is. suppose that we wish to find the debt policy. i.26) for wt .6. but adding government debt.e.e.29) In (2. n ˆ ˆ ˆ b ³ ´ (2.6. That is.30) . at the optimum government debt policy simply transfers wealth from the young to the old (if the debt is positive). the government makes loans to young agents which are financed by taxation. a rate just sufficient to pay the interest and principal on previouslyissued debt. Note that. to get f 0 (k ∗ (b)) − n k (b)(1+n)+b = s f (k (b)) − k (b)f (k (b)) − b. from (2.25). Here. and rt+1 . per capita asset supplies equals savings per capita. Optimal savings for a young agent is st = Ã β (wt − τt ). f 0 (kt+1 ) 1+n 0 Ã Ã ! ! (2. Substituting in (2. from (2.28) Now.27) We can then determine the steady state capital-labor ratio k∗ (b) by setting k∗ (b) = kt = kt+1 in (2.28) we can solve for b ∗ ∗ ∗ 0 ∗ Ã Ã ! ! ˆ = −k(1 + n) + s f (k) − kn. L) = γK α L1−α . then debt b b is negative. If ˆ < 0. u(cy .1 Example Consider the same example as above. i. co ) = ln cy + β ln co . 1+β ! (2. 2. note that ˆ may be positive or negative.11). τt = 0 in the steady state with b = ˆ so that the size of the government debt increases at b. τt . determined by b. Now. the debt increases at the rate n. given that b b) 0 ˆ ˆ as follows: f (k) = n.29). GOVERNMENT DEBT 33 that is. we get f 0 (kt ) − n kt+1 (1+n)+b = s f (kt ) − kt f (kt ) − b. from (2. or from the old to the young (if the debt is negative). where β > 0. we want to find ˆ such that k ∗ (ˆ = k. and 0 < α < 1.27). and F (K.2. which is equal to the interest rate. f 0 (k∗ (b)) 1+n (2. γ > 0. which yields a competitive equilibrium steady state which is socially ˆ optimal.

the equilibrium sequence {kt }∞ is determined by t=0 kt+1 (1 + n) + b = à β 1+β !" !" (1 − α α)γkt α−1 (αγkt − n)b − .7 References Diamond. government debt matters. and β. Ricardian equivalence does not hold here. which can accomplish the same thing in this model. (2.” American Economic Review 55. the optimal quantity of per-capita debt is β αγ 1−α αγ ˆ = (1 − α)γ − b 1+β n n # µ ¶ α " αγ 1−α β(1 − α) α − . in general.30).17). b 2.6. from (2. like social security.34CHAPTER 2. and b ˆ > 0 for α sufficiently small. ˆ < 0 for α sufficiently large. GROWTH WITH OVERLAPPING GENERATIONS Then. That is. . from (2. k∗ (b). 1126-1150. (2.2 Discussion The competitive equilibrium here is in general suboptimal for reasons discussed above. 2. “National Debt in a Neoclassical Growth Model. However. P. = γ n 1+β n à µ ¶ α µ ¶ 1 1−α (1 + n) Here note that. 1965. note that there are other intergenerational transfer mechanisms. 1+n # and the steady state capital-labor ratio. Government debt policy is a means for executing the intergenerational transfers that are required to achieve optimality. given γ. because the taxes required to pay off the currently-issued debt are not levied on the agents who receive the current tax benefits from a higher level of debt today. n. is the solution to k∗ (b)(1 + n) + b = à β 1+β ! (αγ (k∗ (b))α−1 − n)b ∗ (1 − α)γ (k (b)) − 1+n α # Then. But for those same reasons.27) and (2.29).16).

Models of Monetary Economies. Chapter 3. . Kareken. S. O. J. MN. 1980. 1989. Lectures on Macroeconomics.2. REFERENCES 35 Blanchard. and Fischer. Minneapolis.7. and Wallace. Federal Reserve Bank of Minneapolis. N.

36CHAPTER 2. GROWTH WITH OVERLAPPING GENERATIONS .

Cass (1965) and Koopmans (1965) developed the first optimizing models of economic growth. often called “optimal growth” models.e. particularly that of Solow (1956). Later. make statements about welfare). as the model will not exhibit long-run growth. This class of optimal growth models led to the development of stochastic growth models (Brock and Mirman 1972) which in turn were the basis for real business cycle models. which are useful in solving many dynamic models. That is. we will present a simple growth model which illustrates some of the important characteristics of this class of models. 37 . Optimal growth models have much the same long run implications as Solow’s growth model. with the added benefit that optimizing behavior permits us to use these models to draw normative conclusions (i. as they are usually solved as an optimal growth path chosen by a social planner. One objective of this chapter will be to introduce and illustrate the use of discrete-time dynamic programming methods.Chapter 3 Neoclassical Growth and Dynamic Programming Early work on growth theory. Here. “Growth model” will be something of a misnomer in this case. these were theories of growth and capital accumulation in which consumers were assumed to simply save a constant fraction of their income. was carried out using models with essentially no intertemporal optimizing behavior.

The capital stock obeys the law of motion kt+1 = (1 − δ)kt + it . and Technology ∞ X t β u(c t=0 There is a representative infinitely-lived consumer with preferences given by t ). and ct is consumption.4) 3. with 0 ≤ δ ≤ 1 and k0 is the initial capital stock. Assume that limc→0 u0 (c) = ∞. One specification is to endow consumers with . The resource constraints for the economy are ct + it ≤ yt .1 Preferences. limk→0 F1 (k. n) = 0. and strictly quasiconcave. Assume that F (0. NEOCLASSICAL GROWTH AND DYNAMIC PROGRAMMING 3. kt is the capital input. (3.2 Social Planner’s Problem There are several ways to specify the organization of markets and production in this economy. and bounded. ·) is continuously differentiable.3) and nt ≤ 1. all of which will give the same competitive equilibrium allocation. strictly increasing.2) where it is investment and δ is the depreciation rate. and limk→∞ F1 (k. Each period. (3. 1) = 0. where 0 < β < 1.38CHAPTER 3. 1) = ∞. The period utility function u(·) is continuously differentiable. The production technology is given by yt = F (kt . and nt is the labor input.1) where yt is output. homogeneous of degree one. The production function F (·. nt ). (3. which can be supplied as labor. (3. which is given. the consumer is endowed with one unit of time. Endowments. strictly increasing in both arguments. strictly concave.

it is a standard result that the competitive equilibrium is unique and that the first and second welfare theorems hold here. as follows.6) (3.7) does not hold with equality. . Given this.. since u(c) is strictly increasing in c. Therefore.5) (3. As there is no disutility from labor... (3. it is natural to think of kt as a “state . (3.2) to substitute for yt to get (3. we have used (3.kt+1 }t=0 β t u(ct ) subject to ct + kt+1 = f (kt ) + (1 − δ)kt . . nt ≤ 1. k0 given. the quantity of capital available at the beginning of the period. and k0 given.2.5) using (3. the problem can be reformulated as max∞ ∞ X t=0 {ct .5) will be satisfied with equality. Now. suppose that we solve the optimization problem sequentially.3. That is. nt ). This problem appears formidable. which is max ∞ X t=0 {ct . 2. then nt and ct could be increased..7) will hold with equality at the optimum. Firms then purchase capital inputs (labor and capital services) from consumers in competitive markets each period and maximize per-period profits. We can then solve for the competitive equilibrium quantities by solving the social planner’s problem. Here.1) and (3. and increasing utility.nt .. At the beginning of any period t. as nt = 1 for all t. t = 0. (3. 1. SOCIAL PLANNER’S PROBLEM 39 the initial capital stock.it . substitute for it in (3. 1. 1).. and define f (k) ≡ F (k. 2.5). particularly as the choice set is infinite-dimensional. and have them accumulate capital and rent it to firms each period.6). Therefore. the utility that the social planner can deliver to the consumer depends only on kt . kt+1 = (1 − δ)kt + it .kt+1 }∞ t=0 β t u(ct ) subject to ct + it ≤ F (kt . holding constant the path of the capital stock. the competitive equilibrium allocation is the Pareto optimum. Now. if (3. However. Then.7) t = 0.

. and c0 = f (k0 ) + (1 − δ)k0 − g(k0 ). Of course. NEOCLASSICAL GROWTH AND DYNAMIC PROGRAMMING variable” for the problem. say v(k1 ).9) Equation (3.k1 subject to c0 + k1 = f (k0 ) + (1 − δ)k0 . we can write v(k0 ) = max[u(c0 ) + βv(k1 )] c0 .8) is a functional equation or Bellman equation. the choice variables. v(·) is unknown. but the Bellman equation can be used to find it. in period 0 the social planner solves max[u(c0 ) + βv(k1 )] c0 .kt+1 (3. are ct and kt+1 . beginning in period 1. (3.40CHAPTER 3. In period 0. or more generally v(kt ) = max [u(ct ) + βv(kt+1 )] ct . This is a simple constrained optimization problem which in principle can be solved for decision rules k1 = g(k0 ). In most of the cases we will deal with. Since the maximization problem is identical for the social planner in every period. In general. or “control” variables. the Bellman equation satisfies a contraction mapping theorem. if we know the maximum utility that the social planner can deliver to the consumer as a function of k1 . Our primary aim here is to solve for. the optimal decision rules kt+1 = g(kt ) and ct = f (kt ) + (1 − δ)kt − g(kt ). it is straightforward to solve the problem for the first period.8) subject to ct + kt+1 = f (kt ) + (1 − δ)kt .k1 subject to c0 + k1 = f (k0 ) + (1 − δ)k0 . There is a unique function v(·) which satisfies the Bellman equation. That is. or at least to characterize. we cannot solve the above problem unless we know the value function v(·). Within the period. which implies that 1. where g(·) is some function.

guess an initial value function. First. First. and then verify that v(kt ) solves the Bellman equation. Then. That is. Second. then if there are n state variables. .km } = S. if we choose a grid with m values for each state variable. Here. 2. . that is i vj = max[u(c) + βvj−1 ] . The above two implications give us two alternative means of uncovering the value function. 2. i = 2. that i is m values v0 = v0 (ki ). k2 .k subject to c + k0 = f (k) + (1 − δ)k. 1. i.3. .. m. Next. iterate on these values. determining the value function at the j th iteration from the Bellman equation. the accuracy of the approximation depends on how fine the grid is. .8). then we can simply plug v(kt+1 ) into the right side of (3. implication 2 above is useful for doing numerical work. allow the capital stock to take on only a finite number of values. This problem of computational burden as n gets large is sometimes referred to as the curse of dimensionality. One approach is to find an approximation to the value function in the following manner. form a grid for the capital stock... if we are fortunate enough to correctly guess the value function v(·). .c subject to Iteration occurs until the value function converges.2.. i = 1. c + k = f (ki ) + (1 − δ)ki . k ∈ {k1 . This procedure is not too computationally burdensome in this case.. if ki − ki−1 = γ.. the computational burden increases exponentially as we add state variables. then. where m is finite and ki < ki+1 . in particular those which are amenable to judicious guessing... This procedure only works in a few cases. where we have only one state variable. SOCIAL PLANNER’S PROBLEM 41 2.e. given implication 1 above. the search for a maximum on the right side of the Bellman equation occurs over mn grid points.m.. for i = 0. However. then the approximation gets better the smaller is γ and the larger is m. If we begin with any initial function v0 (k) and define vi+1 (k) by vi+1 (k) = max[u(c) + βvi (k0 )] 0 c.. For example. limi→∞ vi+1 (k) = v(k).

e. Then. Next. we can solve (3.16) .12) Now. NEOCLASSICAL GROWTH AND DYNAMIC PROGRAMMING 3.11) on the left and right sides of (3. which gives α βBkt . and δ = 1 t (i. kt+1 (3.12). substituting for the constraint. we can write the Bellman equation as α v(kt ) = max[ln(kt − kt+1 ) + βv(kt+1 )] kt+1 (3. (3.13) kt+1 = 1 + βB and substituting for the optimal kt+1 in (3.1 Example of “Guess and Verify” α Suppose that F (kt . in the objective function on the right side of (3. substitute using (3.9).10) to get α A + B ln kt = max[ln(kt − kt+1 ) + β(A + B ln kt+1 )]. (3. nt ) = kt n1−α .2. 100% depreciation).10) Now.17) (3.42CHAPTER 3. and collecting terms yields A + B ln kt = βB ln βB − (1 + βB) ln(1 + βB) + βA +(1 + βB)α ln kt .14) to get two equations determining A and B: A = βB ln βB − (1 + βB) ln(1 + βB) + βA B = (1 + βB)α Here.11) where A and B are undetermined constants. solve the optimization problem on the right side of (3. 0 < α < 1.15) (3.12) using (3.8). 1 − αβ (3.14) We can now equate coefficients on either side of (3. guess that the value function takes the form v(kt ) = A + B ln kt . (3.16) for B to get B= α .13). (3. u(ct ) = ln ct .

3.18) and solving for k∗ to get k ∗ = (αβ) 1−α . and kt decreasing if k0 > k∗ . with kt increasing if k0 < k∗ . i. Given (3. consumption. but settles down to a steady state where the capital stock. 1 3.19) . k∗ . and output are constant. The steady state for the capital stock. Steady state consumption is c∗ = (1 − αβ)(k ∗ )α .18) α Since ct = kt − kt+1 . we have v(kt ) = max{u[f (kt ) + (1 − δ)kt − kt+1 ] + βv(kt+1 )} kt+1 (3. Figure 3. Next.1.2. Substituting in the objective function for ct using in the constraint. α kt+1 = αβkt . SOCIAL PLANNER’S PROBLEM 43 Then. though we only need B to determine the optimal decision rules.18) gives a law of motion for the capital stock. we have verified that our guess concerning the form of the value function is correct. is determined by substituting kt = kt+1 = k ∗ in (3. substitute for B in (3.13) using (3. we can use (3.1 shows a dynamic path for kt where the initial capital stock is lower than the steady state.17) to get the optimal decision rule for kt+1 .2 Characterization of Solutions When the Value Function is Differentiable Benveniste and Scheinkman (1979) establish conditions under which the value function is differentiable in dynamic programming problems. we can characterize the solution to the social planner’s problem using first-order conditions. consumption and investment (which is equal to kt+1 given 100% depreciation) are each constant fractions of output.e. At this point. Equation (3. that kt converges monotonically to k∗ . That is. a first-order nonlinear difference equation in kt . (3.15) to solve for A.2. and steady state output is y ∗ = (k∗ )α . This economy does not exhibit long-run growth.8). shown in Figure 3. we have α ct = (1 − αβ)kt .18). Supposing that the value function is differentiable and concave in (3. we can show algebraically and in Figure 1.

1: Steady State and Dynamics kt+1 k* 45o αβktα k0 k1 k2 k* kt .Figure 3.

is −u0 [f (kt ) + (1 − δ)kt − kt+1 ] + βv0 (kt+1 ) = 0. v 0 (kt+1 ) = u0 [f (kt+1 ) + (1 − δ)kt+1 − kt+2 ][f 0 (kt+1 ) + 1 − δ]. after substituting using the constraint in the objective function. Therefore. one plus the net marginal product of capital is equal to the inverse of the discount factor. we do not know v0 (·). from investing the foregone consumption in capital. the net benefit must be zero. We can use (3. NEOCLASSICAL GROWTH AND DYNAMIC PROGRAMMING Then.21) to get −u0 [f (kt ) + (1 − δ)kt − kt+1 ] +βu [f (kt+1 ) + (1 − δ)kt+1 − kt+2 ][f 0 (kt+1 ) + 1 − δ] = 0.22) to solve for the steady state capital stock by setting kt = kt+1 = kt+2 = k∗ to get f 0 (k ∗ ) = 1 − 1 + δ. Now. The first term is the benefit. the first-order condition for the optimization problem on the right side of (3. At the optimum. β (3. 0 (3.21) (3. and the second term is the benefit obtained in period t + 1. to the consumer of consuming one unit less of the consumption good in period t. discounted to period t. or.8). at the margin. without knowing v(·). updating one period.23) i. (3.19) we can differentiate on both sides of the Bellman equation with respect to kt and apply the envelope theorem to obtain v 0 (kt ) = u0 [f (kt ) + (1 − δ)kt − kt+1 ][f 0 (kt ) + 1 − δ]. However. .44CHAPTER 3. substitute in (3. from (3.20) The problem here is that. the steady state capital stock depends only on the discount factor and the depreciation rate.20) for v 0 (kt+1 ) using (3.22) or −u0 (ct ) + βu0 (ct+1 )[f 0 (kt+1 ) + 1 − δ] = 0.e.

prices will be such that the consumer will always choose kt+1 > 0).3. While the most straightforward way to determine competitive equilibrium quantities in this dynamic model is to solve the social planner’s problem to find the Pareto optimum. 2. then we can reformulate the consumer’s problem as max ∞ X t β u(w t=0 t {kt+1 }∞ t=0 + rt kt + (1 − δ)kt − kt+1 ) subject to kt ≥ 0 for all t and k0 given. their wealth takes the form of capital). (3. Consumer’s Problem Consumers store capital and invest (i. where wt is the wage rate and rt is the rental rate on capital.. The consumer then solves the following intertemporal optimization problem..24) t = 0. Ignoring the nonnegativity constraints on capital (in equilibrium.24). max∞ ∞ X t=0 {ct .2. . k0 given.3 Competitive Equilibrium Here. SOCIAL PLANNER’S PROBLEM 45 3.kt+1 }t=0 β t u(ct ) subject to ct + kt+1 = wt + rt kt + (1 − δ)kt . and each period they rent capital to firms and sell labor. If we simply substitute in the objective function using (3. Labor supply will be 1 no matter what the wage rate. as consumers receive no disutility from labor.e.. I will simply assert that the there is a unique Pareto optimum that is also the competitive equilibrium in this model.25) . to determine equilibrium prices we need some information from the solutions to the consumer’s and firm’s optimization problems. 1. the first-order conditions for an optimum are −β t u0 (wt + rt kt + (1 − δ)kt − kt+1 ) +β t+1 u0 (wt+1 + rt+1 kt+1 + (1 − δ)kt+1 − kt+2 )(rt+1 + 1 − δ) = 0 (3.2.

To obtain the capital rental rate. Firm’s Problem The firm simply maximizes profits each period. F2 (kt .e. and simplifying. where η is the rate of time preference. we have 1 + r − δ = β . i.28) to get F2 (kt . rt }∞ . either (3.8) allows a solution for the competitive equilibrium sequence of capital stocks {kt }∞ given k0 . nt ) = rt . Note that rt − δ = f 0 (kt ) − δ is the real interest rate and that. NEOCLASSICAL GROWTH AND DYNAMIC PROGRAMMING Using (3.26) that is. which is determined from the dynamic programming problem (3. if we let 1 β = 1+η .23)]. it solves max[F (kt . nt ) = wt . or. 1) = wt . Now. The prices we need to solve for are {wt .28) . then r − δ = η.46CHAPTER 3. 0 (c ) u t 1 + rt+1 − δ (3. We can t=1 then solve for {ct }∞ using (3.nt and the first-order conditions for a maximum are F1 (kt . i. the sequence of factor prices. plug equilibrium quantities into (3. the real interest rate is equal to the rate of time preference. we get βu0 (ct+1 ) 1 = .27) (3.24) to substitute in (3. To solve for t=0 the real wage. (3. kt .26) or (3. nt ) − wt nt − rt kt ].e.25). one plus the interest rate).27) can be used.26) or (3. the intertemporal marginal rate of substitution is equal to the inverse of one plus the net rate of return on capital (i.9). it is straightforward to solve t=0 for competitive equilibrium prices from the first-order conditions for the firm’s and consumer’s optimization problems. Competitive Equilibrium Prices The optimal decision rule. kt+1 = g(kt ).e. in the 1 steady state [from (3.

” in The Econometric Approach to Development Planning. J. Brock. Judd. 233-240. “On the Concept of Optimal Growth. Cambridge. 1998. L. MA. REFERENCES 47 Note that. T. “Optimum Growth in an Aggregative Model of Capital Accumulation. Sargent.” Econometrica 47. W. In an economy with uncertainty. and Ljungvist. Jones. That is.3. Koopmans. Benveniste. MA. R. MIT Press.” Journal of Economic Theory 4. Introduction to Economic Growth. 1972. 1987. 1995. MIT Press. K. Norton. and Sala-I-Martin. Cambridge. McGraw-Hill. rt }∞ . Chicago. 3. X. 1979. a rational expectations equilibrium has the property that consumers and firms may make errors.3. RandMcNally. T. Harvard University Press. 1965. Cass. she knows the whole sequence of prices {wt . L. . 479-513. MA. 727-732. and Scheinkman. “Optimal Economic Growth and Uncertainty: The Discounted Case. C. when the consumer solves her optimization problem. and in equilibrium the forecasts are correct. D.3 References Barro. and Mirman. Recursive Macroeconomic Theory. but those errors are not systematic. 1998.” Review of Economic Studies 32. Cambridge. 2000. Dynamic Macroeconomic Theory. this a “rational t=0 expectations” or “perfect foresight” equilibrium where each period the consumer makes forecasts of future prices and optimizes based on those forecasts. 1965. “On the Differentiability of the Value Function in Dynamic Models of Economics. L. Sargent. T. Numerical Methods in Economics. Economic Growth.

and Woodford. J. . North Holland.. Handbook of Macroeconomics. eds. 1999.48CHAPTER 3. NEOCLASSICAL GROWTH AND DYNAMIC PROGRAMMING Taylor. M..

There exist persistent differences in per capita income across countries. Some basic facts of economic growth (as much as we can tell from the short history in available data) are the following: 1. There are persistent differences in growth rates of per capita income across countries. so that the dynamic programming methods introduced in Chapter 2 can be applied (Lucas’s models are in continuous time). we use discrete-time models. The correlation between the growth rate of income and the level of income across countries is low. Here. we need to have growth models which can successfully confront the data. that the welfare gains from government policies which smooth out business cycle fluctuations are very small compared to the gains from growth-enhancing policies. Macroeconomists are ultimately interested in economic growth because the welfare consequences of government policies affecting growth rates of GDP are potentially very large. 3. as in Lucas (1987). there is stability over time in growth rates 49 . 2.Chapter 4 Endogenous Growth This chapter considers a class of endogenous growth models closely related to the ones in Lucas (1988). Before we can hope to evaluate the efficacy of government policy in a growth context. Among rich countries. In fact. one might argue. 4.

where At = (1 + a)t A0 . Each household member has one unit of time in each period when they are alive. γ t=0 t (4. (4. ct is per capita consumption. and we ask whether this model can successfully explain the above growth facts. there is a dynastic household which gives equal weight to the discounted utility of each member of the household at each date. and it offers some insights with regard to the growth process. The production technology is given by Yt = Ktα (Nt At )1−α .50 CHAPTER 4.1) where 0 < β < 1. growth is unstable. we consider a class of endogenous growth models.2) n constant and N0 given. and there is a wide diversity in growth experience. This neoclassical growth model can successfully account for growth experience in the United States. ENDOGENOUS GROWTH of per capita income. 5.4) . Kt is the aggregate capital stock. and Nt is population. where Nt = (1 + n)t N0 . Next. and At is a labor-augmenting technology factor. which is supplied inelastically as labor. and show that these models can potentially do a better job of explaining the facts of economic growth. and there is little diversity across countries in growth rates. Among poor countries.3) where Yt is aggregate output. γ < 1. but it does very poorly in accounting for the pattern of growth among countries. (4.1 A Neoclassical Growth Model (Exogenous Growth) ∞ X The representative household has preferences given by cγ β Nt t . That is. Here. (4. 4. we first construct a version of the optimal growth model in Chapter 2 with exogenous growth in population and in technology.

5). K0 .kt+1 }t=0 [β(1 + n)(1 + a)γ ]t à γ! xt γ subject to α xt + (1 + n)(1 + a)kt+1 = kt . it is convenient to set up this optimization problem with a change of variables. t ct Also. The resource constraint for this economy is Nt ct + Kt+1 = Yt . Substituting in the objective function for xt using (4. A NEOCLASSICAL GROWTH MODEL (EXOGENOUS GROWTH)51 with a constant and A0 given. use lower case variables to define t quantities normalized by efficiency units of labor.7) " # .6).kt+1 γ " γ # subject to (4.4.2)-(4. that is β(1 + n)(1 + a)γ < 1. and so that we can easily characterize longrun growth paths. That is.4).6). and the initial capital stock. is given. So that we can use dynamic programming methods. for example yt ≡ AYNt . The social planner’s problem is to maximize (4.2)-(4. t = 0. Note here that we require the discount factor for the problem to be less than one. (4. With substitution in (4.1.1) and (4. the social planner’s problem is then max ∞ ∞ X t=0 {xt .5) Note here that there is 100% depreciation of the capital stock each period. the Bellman equation is xt v(kt ) = max + β(1 + n)(1 + a)γ v(kt+1 ) xt . To determine a competitive equilibrium for this economy. for simplicity.5) using (4. 1. 2.. That is. we have α [kt − kt+1 (1 + n)(1 + a)]γ v(kt ) = max + β(1 + n)(1 + a)γ v(kt+1 ) kt+1 γ (4.6) This optimization problem can then be formulated as a dynamic program with state variable kt and choice variables xt and kt+1 . we can solve the social planner’s problem. as the competitive equilibrium and the Pareto optimum are identical. given the value function v(kt ).. let xt ≡ At . .1) subject to (4. We have 0 < α < 1. (4.

we get α−1 −(1 + a)1−γ xγ−1 + βαkt+1 xγ−1 = 0. (4.12) α Also.11) (1 + a)1−γ Then.8) (4.52 CHAPTER 4.10) Now. where x∗ and k∗ are constants. t and we have the following envelope condition α−1 v0 (kt ) = αkt xγ−1 . since yt = kt . (4. ENDOGENOUS GROWTH The first-order condition for the optimization problem on the right side of (4. then on the balanced growth path the level of output per efficiency unit of labor is βα y = (k ) = (1 + a)1−γ ∗ ∗ α " # α 1−α .9) in (4. that is " # 1 1−α βα ∗ k = (4. α Yt kt so that. we can use this to solve for k∗ .7) is −(1 + n)(1 + a)xγ−1 + β(1 + n)(1 + a)γ v0 (kt+1 ) = 0. Since (4. t (4. we will characterize “balanced growth paths.” that is steady states where xt = x∗ and kt = k∗ . on the balanced growth path.13) In addition.8) and simplifying. βα # (4. .6) can be used to solve for x∗ to get βα x = (1 + a)1−γ ∗ " # 1 1−α " (1 + a)1−γ − (1 + n)(1 + a) . the savings rate is st = Kt+1 kt+1 (1 + n)(1 + a) = .10) must hold on a balanced growth path. the savings rate is s∗ = (k∗ )1−α (1 + n)(1 + a). t t+1 (4.9) Using (4.

14). however. long-run growth rates in aggregate variables are determined entirely by exogenous growth in the labor force and exogenous technological change. from (4. produce level effects. which causes the representative household to discount the future at a lower rate. But even though the savings rate is higher in each of these cases. or γ results in an increase in the long-run savings rate.11) we get s∗ = βα(1 + n)(1 + a)γ . Kt .1. α. x∗ . or γ do. using (4. which all look like the one modelled here. This is a counterintuitive result. the increase in β yields increases in the level of output. and that per capita consumption and output grow at the rate a. Therefore. From (4. an increase in β. Note in particular that an increase in any one of α. Therefore. consumption. and growth in per capita income and consumption is determined solely by the rate of technical change. (4. and aggregate output. growth rates remain unaffected. Nt ct . and increases in k∗ and y ∗ . and capital in the long run. Yt . Thus. or γ have no effect on long-run growth. all grow (approximately) at the common rate a + n. Suppose that we consider a number of closed economies. β. β. from (4.13). all countries would not have access to At ). Changes in any of α. Since k∗ . aggregate consumption.11) and (4. as one might anticipate that a country with a high savings rate would tend to grow faster. or γ. We can also show that β(1 + n)(1 + a)γ < 1 implies that an increase in steady state k∗ will result in an increase in steady state x∗ . A NEOCLASSICAL GROWTH MODEL (EXOGENOUS GROWTH)53 Therefore. But if all countries have access .13). Changes in any of the parameters β. Then. we focus on the balanced growth path since it is known that this economy will converge to this path given any initial capital stock K0 > 0. all countries will converge to a balanced growth path where per capita output and consumption grow at the same rate. β.14) Here. results in an increase in the savings rate [from (4. in terms of the logic of the model. and y ∗ are all constant on the balanced growth path. the model tells us that.4.14)]. the differences in the level of per capita income across countries would have to be explained by differences in α. it then follows that the aggregate capital stock. an increase in β leads to an increase in x∗ . For example. given the same technology (and it is hard to argue that.

1 − ut is the time devoted by each agent to human capital accumulation (i. where α > 0. and ut is time devoted by each agent to production. The evidence we have seems to indicate that growth rates and levels of output across countries are not converging. Yt is output. then α cannot vary across countries. and each agent has one unit of time which can be allocated between time in producing consumption goods and time spent in human capital accumulation.e. ENDOGENOUS GROWTH to the same technology. Here. While neoclassical growth models were used successfully to account for long run growth patterns in the United States. Human capital is produced using the technology ht+1 = δht (1 − ut ). Here. for example yt ≡ Ntt .2 A Simple Endogenous Growth Model In attempting to build a model which can account for the principal facts concerning growth experience across countries. The social planner’s problem can then . the above analysis indicates that they are not useful for accounting for growth experience across countries.54 CHAPTER 4. to permit economic factors to determine long-run growth rates. preferences are as in (4. and introduce physical capital in the next section. to focus on the essential mechanism at work. we will use lower case letters to denote variables in per capita Y terms.15) where δ > 0. education and acquisition of skills). in contrast to what the model predicts. (4. ht is the human capital possessed by each agent at time t. We will construct a model which abstracts from physical capital accumulation. and this leaves an explanation of differences in income levels due to differences in preferences. 4.1). One way to do this is to introduce human capital accumulation. The production technology is given by Yt = αht ut Nt . it would seem necessary to incorporate an endogenous growth mechanism. and h0 is given. the production function is linear in quality-adjusted labor input. That is. This seems like no explanation at all.

. from (4.15) and (4. ht+1 . That is. . then ut = 0.16) and (4.. this could not be optimal. γ " γ ct . Again.16). t + 2. (4.2. and ct = 0 from (4. A SIMPLE ENDOGENOUS GROWTH MODEL 55 be formulated as a dynamic programming problem.15) and (4. so we must have ∂L = λt αht − μt δht = 0 ∂ut at the optimum. then ut = 1. where the state variable is ht and the choice variables are ct . Therefore ∂ut ≤ 0.16). since the marginal utility of consumption goes to infinity as consumption goes to zero.19) . In addition. the Bellman equation for the social planner’s problem is v(ht ) = max subject to ct = αht ut (4. the first derivative of the Lagrangian with respect to ut is ∂L = λt αht − μt δht ∂ut ∂L Now. ∂ht+1 (4.17) (4.16).18) (4. Then.4. But then. If ∂ut < 0. this could ∂L ∂L not be an optimal path. .15). we have hs = cs = 0 for s = t + 1. t ∂ct ∂L = β(1 + n)v0 (ht+1 ) − μt = 0. the Lagrangian for the optimization problem on the right side of the Bellman equation is L= cγ t + β(1 + n)v(ht+1 ) + λt (αht ut − ct ) + μt [δht (1 − ut ) − ht+1 ]. and ut .ht+1 ct + β(1 + n)v(ht+1 ) γ # where λt and μt are Lagrange multipliers.ut . But. if ∂ut > 0. Two first-order conditions for an optimum are then ∂L = cγ−1 − λt = 0.

ht 1 [β(1 + n)δ γ ] 1−γ ut = 1 − ut (4. ENDOGENOUS GROWTH We have the following envelope condition: v0 (ht ) = αut λt + λt α(1 − ut ). a constant. a condition we will assume holds. using (4. from (4.22) Then. and (4. Any path {ut }∞ satisfying (4.17)-(4.1 for the case where [β(1 + n)]1−γ δ −γ < 1. as the representative consumer would be spending all available time accumulating human capital which is never used to produce in the future. t t+1 (4. for all t) has the property that limt→∞ ut = 0.23).15). or. using (4.20) Therefore.16).56 CHAPTER 4.15). (4. we can rewrite (4. substituting in (4. Thus the only solution. is ut = u = 1 − [β(1 + n)δ γ ] 1−γ for all t.21) (4.23) . we then get β(1 + n)δcγ−1 − cγ−1 = 0. ut 1 Now. we get 1 ht+1 = [β(1 + n)δ] 1−γ . Therefore.23) which is not stationary (a t=0 stationary path is ut = u.20). (4.17). we obtain: [β(1 + n)δ] 1−γ = or ut+1 1 δ(1 − ut )ut+1 .23) is a first-order difference equation in ut depicted in Figure 4. which cannot be an optimum.22). ct (4.21) as an equation determining the equilibrium growth rate of consumption: 1 ct+1 = [β(1 + n)δ] 1−γ . v0 (ht ) = αcγ−1 t From (4.

1: Determination of equilibrium ut ut+1 45o u 0 u 1 ut .Figure 4.

except that the production technology is given by Yt = Ktα (Nt ht ut )1−α . then growth rates are positive. since growth rates are constant from for all t.3. Second.3 Endogenous Growth With Physical Capital and Human Capital The approach here follows closely the model in Lucas (1988). Therefore. The model is identical to the one in the previous section. educational policy. the level of per capita income (equal to per capita consumption here) is dependent on initial conditions. the level of income is determined by h0 . 1 If [β(1 + n)δ] 1−γ > 1 (which will hold for δ sufficiently large). except that we omit his treatment of human capital externalities.4. this economy can exhibit unbounded growth. Growth rates depend in particular on the discount factor (growth increases if the future is discounted at a lower rate) and δ. the initial stock of human capital. ENDOGENOUS GROWTH WITH PHYSICAL CAPITAL AND HUMAN CAPITAL57 and human capital grows at the same rate as consumption per capita. the economy grows at a higher rate). 4. and the economy’s resource constraint is Nt ct + Kt+1 = Ktα (Nt ht ut )1−α . countries which are initially relatively rich (poor) will tend to stay relatively rich (poor). The fact that other factors besides exogenous technological change can affect growth rates in this type of model opens up the possibility that differences in growth across countries could be explained (in more complicated models) by factors including tax policy. That is. which is a technology parameter in the human capital accumulation function (if more human capital is produced for given inputs. There are two important results here. The lack of convergence of levels of income across countries which this model predicts is consistent with the data. where Kt is physical capital and 0 < α < 1. and savings behavior. even if there is no growth in population (n = 0) and given no technological change. The first is that equilibrium growth rates depend on more than the growth rates of exogenous factors. Here.

28) respectively.29).30) and (4. ct .26) (4. ENDOGENOUS GROWTH As previously. In the dynamic program associated with the social planner’s optimization problem. ht+1 ) γ # (4. ht ) = subject to α ct + (1 + n)kt+1 = kt (ht ut )1−α ct . and four choice variables. ht+1 )+λt [kt (ht ut )1−α −ct −(1+n)kt+1 ]+μt [δht (1−ut )−ht+1 ] γ The first-order conditions for an optimum are then ∂L = cγ−1 − λt = 0. t ∂ct ∂L α = λt (1 − α)kt h1−α u−α − μt δht = 0. ht+1 ) = 0.26) and (4.28) and (4. ht ) = λt (1 − α)kt h−α u1−α + μt δ(1 − ut ) t t (4.32) . The Bellman equation for this dynamic program is v(kt .31) Next. then use (4.ut . and kt+1 . After simplifying. t t+1 (4.58 CHAPTER 4.27) (4.24) (4. kt and ht .kt+1 .30) (4. ht+1 . We also have the following envelope conditions: α−1 v1 (kt .31) to substitute in (4. ut .25).29) and (4.29) ∂kt+1 (4. we use lower case letters to denote per capita quantities. (4. ht+1 ) − μt = 0.25) The Lagrangian for the constrained optimization problem on the right side of the Bellman equation is then cγ α L = t +β(1+n)v(kt+1 . we obtain the following two equations: α−1 −cγ−1 + βcγ−1 αkt+1 (ht+1 ut+1 )1−α = 0. ∂ht+1 (4. there are two state variables.28) ht+1 = δht (1 − ut ) ∂L = −λt (1 + n) + β(1 + n)v1 (kt+1 .24) and (4. t t ∂ut ∂L = β(1 + n)v2 (kt+1 . use (4. ht ) = λt αkt (ht ut )1−α α v2 (kt .ht+1 max " γ ct + β(1 + n)v(kt+1 .27) to substitute for λt and μt in (4.

34) (4. from (4. we have ct kt+1 α−1 + (1 + n) = kt (ht ut )1−α . (4. along which physical capital.3. and per capita consumption all grow at the same rate along the balanced growth path. kt kt Then. we get (1 + μc )1−γ α−1 = kt (ht ut )1−α βα Equations (4.36) 1 + μc = 1 + μk = 1 + μh = 1 + μ = [δβ(1 + n)] 1−γ .36). which implies that 1 + μh . human capital. human capital. which implies t that μc = μk . and consumption. since ut is a constant. Let μk . (4. ut+1 . dividing (4. human capital. Next. μh .37) .24) through by kt . i. δ a constant. respectively. on the balanced growth path.32) and backdating by one period. and consumption grow at constant rates.36) then imply that ct (1 + μc )1−γ . Thus per capita physical capital.35) and (4. and simplifying. Therefore.25). we wish to use (4.24).35) (4. along the balanced growth path. ENDOGENOUS GROWTH WITH PHYSICAL CAPITAL AND HUMAN CAPITAL59 α α −cγ−1 kt h−α u−α + δβ(1 + n)cγ−1 kt+1 h−α u−α = 0. rearranging (4.33). we get ut = 1 − (1 + μc )1−γ (1 + μk )−α (1 + μh )α = δβ(1 + n).33) Now. 1 (4. and growth rates in (4. (4. it must be the case that μk = μh .4. t t+1 t t t+1 t+1 (4. From (4. we then have 1 + μh = δ(1 − ut ).33) to characterize a balanced growth path.34).25). and we can determine this common rate from (4. + (1 + n)(1 + μk ) = kt βα c But then kt is a constant on the balanced growth path. substituting for ut .32).e. Also. and μc denote the growth rates of physical capital. and (4.

M. J. and Woodford. “On the Mechanics of Economic Development. 1988.. 3-42. R.38) In general then.60 CHAPTER 4. eds. New York. ENDOGENOUS GROWTH Note that the growth rate on the balanced growth path in this model is identical to what it was in the model of the previous section. Models of Business Cycles. from (4. S.. and Prescott.4 References Lucas. or δ) also cause the growth rate of per capita consumption and income to increase. factors which cause the savings rate to increase (increases in β. Handbook of Macroeconomics.36) and (4. Parente. n. .37) and (4.38).” Journal of Monetary Economics 22. Basil Blackwell. The savings rate in this model is st = Kt+1 kt+1 (1 + n) = α−1 Yt kt kt (ht ut )1−α Using (4. Taylor. North Holland.37). 1999. 4. E.E. MIT Press. 2000. on the balanced growth path we then get st = α [δ γ β(1 + n)] 1−γ 1 (4.E. 1987. Lucas. Barriers to Riches. R.

For example. 5. where 0 < pi < 1. We will first provide an outline of expected utility theory. Now suppose two lotteries over consumption. if there is uncertainty. we describe consumer preferences in terms of the ranking of consumption bundles.Chapter 5 Choice Under Uncertainty In this chapter we will introduce the most commonly used approach to the study of choice under uncertainty.1 Expected Utility Theory In a deterministic world. where a consumer’s preferences over certain quantities of consumption goods are described by the function u(c). Expected utility maximization by economic agents permits the use of stochastic dynamic programming methods in solving for competitive equilibria. i = 1. then preferences are defined in terms of how consumers rank lotteries over consumption bundles. 61 . suppose a world with a single consumption good. The axioms of expected utility theory imply a ranking of lotteries in terms of the expected value of utility they yield for the consumer. Lottery i gives the consumer c1 units of consumption with probability pi . and c2 i i units of consumption with probability 1−pi . However. This stochastic growth model is the basis for real business cycle theory. expected utility theory. and then illustrate the use of stochastic dynamic programming in a neoclassical growth model with random disturbances to technology. 2. where c is consumption.

1.1). This states that the consumer prefers the expected value of the lottery with certainty to the lottery itself. If u(c) is strictly concave. u(E[c])) (see Figure 5. that is E[u(c)] ≤ u (E[c]) . (5. CHOICE UNDER UNCERTAINTY Then. as in Figure 5. where α and β are constants.1) where E is the expectation operator. we have. 1 1 2 2 would strictly prefer lottery 2 to lottery 1 if p1 u(c1 ) + (1 − p1 )u(c2 ) < p2 u(c1 ) + (1 − p2 )u(c2 ).1) holds as a strict inequality.1) holds with equality. That is.62 CHAPTER 5. i i and the consumer would strictly prefer lottery 1 to lottery 2 if p1 u(c1 ) + (1 − p1 )u(c2 ) > p2 u(c1 ) + (1 − p2 )u(c2 ). 1 1 2 2 Many aspects of observed behavior toward risk (for example. since u(c) is strictly concave. Now. this implies Jensen’s inequality.2) .4) (5.3) (5. a risk averse consumer would pay to avoid risk. ¯ then clearly (5. An expected utility maximizing consumer will be risk averse with respect to all consumption lotteries if the utility function is strictly concave. the expected utility the consumer receives from lottery i is pi u(c1 ) + (1 − pi )u(c2 ). the observation that consumers buy insurance) is consistent with risk aversion. That is. α + βc ≥ u(c). we can show that (5. with certainty. (5. If the consumer receives constant consumption. c. take a tangent to the function u(c) at the point (E[c]. 1 1 2 2 and would be indifferent if p1 u(c1 ) + (1 − p1 )u(c2 ) = p2 u(c1 ) + (1 − p2 )u(c2 ). This tangent is described by the function g(c) = α + βc. In the case where consumption is random. and we have α + βE[c] = u(E[c]).

Figure 5.1: Jensen’s Inequality α + βc u(c) u[E(c)] E(c) c .

c2 − c1 c2 − c1 A point on the line AB denotes the expected utility the agent receives for a particular value of p. .3). or. EXPECTED UTILITY THEORY 63 for c ≥ 0. u(E[c]) > E[u(c)]. In this case. many individuals engage in lotteries with small stakes where the expected payoff is negative.2.1 Anomalies in Observed Behavior Towards Risk While expected utility maximization and a strictly concave utility function are consistent with the observation that people buy insurance. Jensen’s inequality is reflected in the fact that AB lies below the function u(c). and B implies p = 1. using (5. consider a consumption lottery which yields c1 units of consumption with probability p and c2 units with probability 1 − p.1) takes the form pu(c1 ) + (1 − p)u(c2 ) < u (pc1 + (1 − p)c2 ) . Note that the distance DE is the disutility associated with risk. (5. and that this distance will increase as we introduce more curvature in the utility function.1. with strict inequality if c 6= E[c]. where 0 < p < 1 and c2 > c1 . as the consumer becomes more risk averse.5. i. As an example. for example p = 0 yields expected utility u(c1 ) or point A.1. Since the expectation operator is a linear operator. and given that c is random we have α + βE[c] > E[u(c)]. The line AB is given by the function u(c2 ) − u(c1 ) c2 u(c1 ) − c1 u(c2 ) f (c) = + c. the difference u (pc1 + (1 − p)c2 ) − [pu(c1 ) + (1 − p)u(c2 )] is given by DE. some observed behavior is clearly inconsistent with this. In Figure 5. For example. " # 5. we can take expectations through (5.e.4).

2: Jensen’s Inequality Again B u(c2) u[pc1+(1-p)c2] pu(c1)+(1-p)u(c2) u(c1) A E D u(c) c1 pc1+(1-p)c2 c2 c .Figure 5.

the pricing of risky assets. a preference for lottery 4 over lottery 3 gives . lottery 4 yields $5 million with probability .89u(0) < .6) and clearly (5.9.1u(5) + . lottery 3 yields $1 million with probability . this is still the standard approach used in most economic problems which involve choice under uncertainty.11u(1) + . CHOICE UNDER UNCERTAINTY Another anomaly is the “Allais Paradox.1. as we will show. .89u(1) + .9u(0). Expected utility theory has proved extremely useful in the study of insurance markets.2 Measures of Risk Aversion With expected utility maximization. (5.01u(0).89.01u(0).” Here. if u(·) is an agent’s utility function.1u(5) + .1u(5) + .1.1u(5) + . That is.11u(1) > . Experiments show that most people prefer lottery 1 to lottery 2. That is. Though there appear to be some obvious violations of expected utility theory. then a preference for lottery 1 over lottery 2 gives u(1) > . or . and in modern macroeconomics. $1 million with probability .11u(1) < .6). Similarly. choices made under uncertainty are invariant with respect to affine transformations of the utility function.5) 5. and 0 with probability . suppose a utility function v(c) = α + βu(c). But this is inconsistent with expected utility theory (whether the person is risk averse or not is irrelevant).5) is inconsistent with (5. and they maximize expected utility. and lottery 4 to lottery 3.11 and 0 with probability . which a person can enter at zero cost. suppose that there are four lotteries. or . Lottery 1 involves a payoff of $1 million with certainty.89.9u(0). lottery 2 yields a payoff of $5 million with probability .1 and 0 with probability .64 CHAPTER 5.01. (5.

1−γ . that the measure of absolute risk aversion is twice the maximum amount that the consumer would be willing to pay to avoid one unit of variance for small risks. Here. u0 (c) A utility function which has the property that ARA(c) is constant for all c is u(c) = −e−αc . Then. ARA(c) = − u00 (c) . that we have v00 (c) = βu00 (c). For this function. Any measure of risk aversion should clearly involve u00 (c). i. since risk aversion increases as curvature in the utility function increases. A measure of risk aversion which is invariant to affine transformations is the measure of absolute risk aversion.1. note that for the function v(c). lotteries are ranked in the same manner with v(c) or u(c) as the utility function. However. u0 (c) A utility function for which RRA(c) is constant for all c is u(c) = where γ ≥ 0. αe−αc It can be shown. which have no effect on behavior. α > 0. 65 since the expectation operator is a linear operator. the second derivative is not invariant to affine transformations. through Taylor series expansion arguments. −γc−(1+γ) RRA(c) = −c =γ c−γ c1−γ − 1 . we have ARA(c) = − −α2 e−αc = α. EXPECTED UTILITY THEORY where α and β are constants with β > 0.5. An alternative is the relative risk aversion measure.e. we have E[v(c)] = α + βE[u(c)]. Thus. RRA(c) = −c u00 (c) .

zt . where F (·.66 CHAPTER 5. where β > 0.i. Since u00 (c) = 0 and u0 (c) = β. ct will be random. kt is the capital input. Note here that. homogeneous of degree one. strictly concave. The representative consumer has preferences given by E0 ∞ X t=0 β t u(ct ). and zt is a random technology disturbance. The representative consumer has 1 unit of labor available in each period. which is essentially the stochastic optimal growth model studied by Brock and Mirman (1972). That is. In each period.) random variables (each period zt is an independent draw from a fixed probability distribution G(z)). which is supplied inelastically. 5. A consumer is risk neutral if they have a utility function which is linear in consumption. we have ARA(c) = RRA(c) = 0. The measure of relative risk aversion can be shown to be twice the maximum amount per unit of variance that the consumer would be willing to pay to avoid a lottery if both this maximum amount and the lottery are expressed as proportions of an initial certain level of consumption. and twice differentiable.2 Stochastic Dynamic Programming We will introduce stochastic dynamic programming here by way of an example. the current realization. nt is the labor input. {zt }∞ is t=0 a sequence of independent and identically distributed (i. nt ). so that the consumer cares only about the expected value of consumption. is learned . and increasing in both argument. We then have E[u(c)] = βE[c]. where 0 < β < 1. The production technology is given by yt = zt F (kt . that is u(c) = βc. ·) is strictly quasiconcave. CHOICE UNDER UNCERTAINTY Note that the utility function u(c) = ln(c) has RRA(c) = 1. in general. Here.d. u(·) is strictly increasing. ct is consumption. and E0 is the expectation operator conditional on information at t = 0.

Showing that the competitive equilibrium is Pareto optimal here is a straightforward extension of general equilibrium theory. The second approach is to have spot market trading with rational expectations. At each date. The representative consumer accumulates capital over time by saving. all contingent claims markets (and there are potentially very many of these) clear at t = 0. in period t labor is sold at the wage rate wt and capital is rented at the rate rt . The law of motion for the capital stock is kt+1 = it + (1 − δ)kt . both of which yield the same unique Pareto optimal allocation. The first is to follow the approach of Arrow and Debreu (see Arrow 1983 or Debreu 1983). date T ) conditional on a particular realization of the sequence of technology shocks. and as information is revealed over time. markets clear at every date t for every . the contracts which specify how much labor and capital services are to be delivered at each date are written at date t = 0. where it is investment and δ is the depreciation rate. zT }. with 0 < δ < 1. the consumer rents capital and sells labor at market prices. before decisions are made. In equilibrium. contracts are executed according to the promises made at t = 0. and in each period he/she rents capital and sells labor to the representative firm. That is.2. STOCHASTIC DYNAMIC PROGRAMMING 67 at the beginning of the period.. However.2. z1 . A contingent claim is a promise to deliver a specified number of units of a particular object (in this case labor or capital services) at a particular date (say.1 Competitive Equilibrium In this stochastic economy. {z0 . there are two very different ways in which markets could be organized. with many state-contingent commodities. and makes an optimal savings decision given his/her beliefs about the probability distribution of future prices. the representative firm and the representative consumer get together and trade contingent claims on competitive markets. The resource constraint for this economy is ct + it = yt . 5.5.. z2 . At t = 0.. . In a competitive equilibrium.

where f (k) ≡ F (k. ex ante. agents are not systematically fooled.2 Social Planner’s Problem Since the unique competitive equilibrium is the Pareto optimum for this economy. a rational expectations equilibrium is equivalent to the Arrow Debreu equilibrium.kt+1 subject to ct + kt+1 = zt f (kt ) + (1 − δ)kt . . as complete markets are required for efficient risk sharing. In this representative agent environment. we can simply solve the social planner’s problem to determine competitive equilibrium quantities.2. 1). but this will not be true in models with heterogeneous agents. CHOICE UNDER UNCERTAINTY possible realization of the random shocks {z0 .68 CHAPTER 5. since they make efficient use of available information. However. . In equilibrium.. zt+1 )] ct . given the nature of uncertainty. and zt is given by nature and known when decisions are made concerning the choice variables ct and kt+1 . In the problem. where kt is determined by past decisions. z1 . in the sense that agents’ beliefs about the probability distributions of future prices are the same as the actual probability distributions. the relevant state variables are kt and zt . to be small probability events. In those models. z2 . The social planner’s problem is max E0 ∞ X t β u(c t=0 t) {ct . complete markets in contingent claims are necessary to support Pareto optima as competitive equilibria. The Bellman equation is written as v(kt .. zt }. agents can be surprised in that realizations of zt may occur which may have seemed. In equilibrium expectations are rational.. zt ) = max [u(ct ) + βEt v(kt+1 . 5.kt+1 }∞ t=0 subject to ct + kt+1 = zt f (kt ) + (1 − δ)kt . Setting up the above problem as a dynamic program is a fairly straightforward generalization of discrete dynamic programming with certainty.

nt ) = kt n1−α . substituting for the optimal kt+1 in (5.3 Example α Let F (kt .7) Solving the optimization problem on the right-hand side of the above equation gives βB (5. is unknown.2.5. v(·.8) kt+1 = zt kα . ·) is the value function and Et is the expectation operator conditional on information in period t.. ·). zt ) = A + B ln kt + D ln zt The Bellman equation for the social planner’s problem.e. kt+1 = g(kt . Guess that the value function takes the form v(kt . i. ct is known but ct+i. v (·.. we get α zt kt A + B ln kt + D ln zt = ln 1 + βB Ã ! α βBzt kt + βA + βB ln 1 + βB Ã ! + βDμ (5. What we wish to determine in the above problem are the value function. with 0 < α < 1. in period t. 3.9) . That is. Note that. i = 1. the value of the problem at the beginning of period t + 1 (the expected utility of the representative agent at the beginning of period t + 1) is uncertain as of the beginning of period t. kt+1 (5.. . and t E[ln zt ] = μ. zt ) and ct = zt f (kt ) + (1 − δ)kt − g(kt .7). 5. after substituting for the resource constraint and given that nt = 1 for all t. is then α A+B ln kt +D ln zt = max {ln[zt kt − kt+1 ] + βEt [A + B ln kt+1 + D ln zt+1 ]} . zt ).2. kt+1 or α A+B ln kt +D ln zt = max {ln[zt kt − kt+1 ] + βA + βB ln kt+1 + βDμ} . STOCHASTIC DYNAMIC PROGRAMMING 69 Here. and optimal decision rules for the choice variables. δ = 1. 1 + βB t Then. u(ct ) = ln ct . 2.

Note that the economy will not converge to a steady state here. solving (5. These time series .13) and (5. This model is easy to simulate on the computer. consumption.13) (5.14) Here.10)-(5. (5.12) Then. 1 αβ βμ A= ln(1 − αβ) + ln(αβ) + 1−β 1 − αβ 1 − αβ (5. and investment.9) gives 1 A = ln 1 + βB Ã ! βB + βA + βB ln 1 + βB B = α + αβB D = 1 + βB Ã ! + βDμ (5. consumption. CHOICE UNDER UNCERTAINTY Our guess concerning the value function is verified if there exists a solution for A. B. Substituting for B in (5.12) for A. and investment. B.10) (5. the optimal decision rule for ct is α ct = (1 − αβ)zt kt . there will be convergence to a stochastic steady state.13) and (5. To do this.14) to determine time series for consumption and investment. However.11) (5. some joint probability distribution for output.e.14) determine the behavior of time series for ct and kt (where kt+1 is investment in period t). α and since ct = zt kt − kt+1 . i.8) gives the optimal decision rule α kt+1 = αβzt kt . Equating coefficients on either side of equation (5.70 CHAPTER 5. and then use (5. simply assume some initial k0 . as technology disturbances will cause persistent fluctuations in output. determine a sequence {zt }T using a random t=0 number generator and fixing T. and D. and D gives B= D= " α 1 − αβ 1 1 − αβ # We have now shown that our conjecture concerning the value function is correct.

5. and then to choose parameters based on long-run evidence and econometric studies. For an overview of this literature. L. K. 479-513. Real business cycle (RBC) analysis is essentially an exercise in modifying this basic stochastic growth model to fit the post-war U. NJ. Cambridge. Cooley. if we take logs through (5.S. yt = zt kt . For example. Princeton.A. Vol. 1995.14). and the output matched to the actual data to judge the fit. Harvard University Press.S. time series data. . Arrow. the log of investment is much more variable (about trend) than is the log of output. Brock. Frontiers of Business Cycle Research. 1983.3 References Arrow.3. and the log of output is more variable than the log of consumption. “Optimal Economic Growth and Uncertainty: the Discounted Case. the model is run on the computer. Following that. 1972. MA. Collected Papers of Kenneth J. W. Princeton University Press. T.” Journal of Economic Theory 4. The fitted model can then be used (given that the right amount of detail is included in the model) to analyze the effects of changes in government policies. 2.5. But in the data. General Equilibrium. see Prescott (1986) and Cooley (1995). and Mirman. employment is constant α here while it is variable in the data. The basic approach is to choose functional forms for utility functions and production functions. though there will be obvious ways in which this model does not fit the data. time series.13) and (5. we get ln kt+1 = ln αβ + ln yt and ln ct = ln(1 − αβ) + ln yt We therefore have var(ln kt+1 ) = var(ln ct ) = var(ln yt ). REFERENCES 71 will have properties that look something like the properties of post-war detrended U. given that output. Also.

Recursive Methods in Economic Dynamics. CHOICE UNDER UNCERTAINTY Debreu. “Theory Ahead of Business Cycle Measurement. 922. 1983. MIT Press. eds.” Federal Reserve Bank of Minneapolis Quarterly Review. MA. N. 1999. Cambridge. and Lucas. G. E. 1989. . M.. J. Cambridge University Press. Mathematical Economics: Twenty Papers of Gerard Debreu. Prescott. Fall. Taylor.72 CHAPTER 5. L. Handbook of Macroeconomics. Cambridge. 1986. Stokey. Harvard University Press.. Recursive Macroeconomic Theory. North Holland. and Ljungvist. R. 2000. Sargent. T. and Woodford. Cambridge. MA.

look quite similar. captured in the stochastic Euler equations from the representative consumer’s problem. These two topics are treated together because there is a close relationship between the behavior of consumption and asset prices in this class of models. The stochastic implications of consumption theory and asset pricing theory. 73 . Friedman’s and Modigliani and Brumberg’s ideas can all be exposited in a rigorous way in the context of the class of representative agent models we have been examining. relative to aggregate income. asset pricing theory typically treats aggregate consumption as exogenous while determining equilibrium asset prices. That is.Chapter 6 Consumption and Asset Pricing In this chapter we will examine the theory of consumption behavior and asset pricing in dynamic representative agent models. consumption theory typically treats asset prices as being exogenous and determines optimal consumption-savings decisions for a consumer. Traditional theories of consumption which explain this fact are Friedman’s permanent income hypothesis and the life cycle hypothesis of Modigliani and Brumberg. However.1 Consumption The main feature of the data that consumption theory aims to explain is that aggregate consumption is smooth. 6.

5) . 1+r 1+r and the envelope theorem gives v 0 (At ) = u0 wt + At − µ µ ¶ (6. lim t→∞ (1 + r)t This condition and (6.1) where 0 < β < 1.74 CHAPTER 6.. . Consider a consumer with initial assets A0 and preferences ∞ X t=0 β t u(ct ). where income is exogenous. 1+r ¶ ¸ The first-order condition for the optimization problem on the righthand side of the Bellman equation is 1 0 At+1 − u wt + At − + βv 0 (At+1 ) = 0. strictly concave.2) gives the intertemporal budget constraint for the consumer. At+1 }∞ to maximize (6.2) for t = 0. 2.1. The consumer’s budget constraint is At+1 = (1 + r)(At − ct + wt ). 1+r ¶ (6. 1. with the value function v(At ) assumed to be concave and differentiable. and u(·) is increasing.2). where r is the one-period interest rate (assumed constant over time) and wt is income in period t. ∞ ∞ X X ct wt = A0 + (6.. (6. We also assume the no-Ponzi-scheme condition At = 0. (6.1) t=0 subject to (6. ct is consumption.4) At+1 . and twice differentiable.1 Consumption Behavior Under Certainty The model we introduce here captures the essentials of consumptionsmoothing behavior which are important in explaining why consumption is smoother than income. Formulating this problem as a dynamic program..3) t t t=0 (1 + r) t=0 (1 + r) The consumer’s problem is to choose {ct . the Bellman equation is v(At ) = max u wt + At − At+1 ∙ µ At+1 + βv(At+1 ) . CONSUMPTION AND ASSET PRICING 6.

i. where. in this case u(c) = c1−α − 1 . t t=0 (1 + r) ∞ X ! (6. suppose a period is a quarter. ct (6. the impact on consumption of a temporary increase in income is very small.6. Also.5) and (6.e.7) Here.7) implies that a $1 increase in current income gives an increase in current consumption of $.3). consumption in each period is just a constant fraction of discounted lifetime wealth or “permanent income. note that (6. we get r c= 1+r µ ¶Ã wt A0 + . but the consumer is able to t=0 smooth consumption perfectly by borrowing and lending in a perfect capital market. 1 Now. Then (6.6) we get 1 ct+1 = [β(1 + r)] α . then (6. Now. CONSUMPTION Therefore.7) implies that the response of consumption to an increase in permanent income is very small. That is.8) . the intertemporal marginal rate of substitution is equal to one plus the interest rate at the optimum. If 1 + r = β .” The income stream given by {wt }∞ could be highly variable. from (6. This is an important implication of the permanent income hypothesis: because consumers smooth consumption over time. consider some special cases. from (6. and take r = . if the interest rate is equal to the discount rate.4) using (6.01 (an interest rate of approximately 4% per annum).2) gives u0 (ct ) =1+r βu0 (ct+1 ) 75 (6.6) That is.6) gives ct = ct+1 = c for all t.0099. 1−α where γ > 0. but assumes a particular utility function. substituting in (6.1. Another example permits the discount factor to be different from the interest rate.

2). At+1 1+r and the first-order condition for the maximization problem on the righthand side of the Bellman equation is − 1 0 At+1 u (At + wt − ) + βEt v1 (At+1 . Again. From (6. is a random variable which becomes known at the beginning of period t. where u(·) has the same properties as in the previous section.9) ∙ ¸ We also have the following envelope condition: v1 (At . wt+1 ). but Friedman did not develop his theory in the context of an optimizing model with uncertainty. CONSUMPTION AND ASSET PRICING so that consumption grows at a constant rate for all t. and solving for c0 using (6.8) we have ct = c0 [β(1 + r)] α . wt ) for the consumer’s problem. but now the consumer’s income. the Bellman equation associated with the consumer’s problem is At+1 v(At . the consumption path is smooth. aimed at explaining the regularities in short run and long run consumption behavior.2 Consumption Behavior Under Uncertainty Friedman’s permanent income hypothesis was a stochastic theory.76 CHAPTER 6. we obtain c0 = 1 − β (1 + r) h 1 α 1−α α t i " wt A0 + . and the following is essentially Hall’s model. t t=0 (1 + r) ∞ X # 6. 1+r 1+r (6.3).1. wt+1 ) . wt ) = u0 (At + wt − At+1 ). wt . Consider a consumer with preferences given by E0 X β t u(ct ).10) . This was later done by Hall (1978). wt ) = max u(At + wt − ) + βEt v(At+1 . Given a value function v(At . The consumer’s budget constraint is given by (6. 1+r (6.

asset prices move in such a way as to induce the representative consumer to consume what is produced in the current period.2). without knowing the utility function. However. Essentially. i. and these models fit the properties of aggregate consumption well. (6.11) gives c ¯ 2 β(1 + r) − 1 Et ct = ct . " # . the representative consumer has an incentive to smooth consumption. and (6.e.11) states that u0 (ct ) is a martingale with drift. Basically. we obtain Et u0 (ct+1 ) = 1 u0 (ct ). But in the aggregate. β(1 + r) 77 (6. the interest rates at which consumers can borrow are typically much higher than the interest rates at which they can lend. In a real business cycle model.9). β(1 + r) so that consumption is a martingale with drift. is that capital markets are imperfect in practice. as in Hall’s model) in general equilibrium. That is. (6. for example. consumers respond more strongly to changes in current income than theory predicts they should. CONSUMPTION Therefore. in practice. A second possible explanation. where c > 0 is a constant.11) Here. A large body of empirical work (summarized in Hall 1989) comes to the conclusion that (6. In a real business cycle model. the problem is that consumption is too variable in the data relative to what the theory predicts. from (6. There are at least two explanations for the inability of the permanent income model to fit the data.1. this does not tell us much about the path for consumption.6. If we suppose that u(·) is quadratic. the ability of consumers to smooth consumption is limited by the investment technology. That is. (6. consumption is smooth in the sense that the only information required to predict future consumption is current consumption. The first is that much of the work on testing the permanent income hypothesis is done using aggregate data. That is. which has been explored by many authors (see Hall 1989). (6.10). u(ct ) = − 1 (¯ − ct )2 .11) is a stochastic Euler equation which captures the stochastic implications of the permanent income hypothesis for consumption.11) does not fit the data well. interest rates are not exogenous (or constant.

Output is produced on n productive units.13) but our interest here is in determining competitive equilibrium prices. However. (6. This asset pricing model is sometimes referred to as the ICAPM (intertemporal capital asset pricing model) or the consumption-based capital asset pricing model. and zit the . developed by Lucas (1978).12) where 0 < β < 1 and u(·) is strictly increasing. Here yit is random.2 Asset Pricing In this section we will study a model of asset prices. strictly concave. Each period. It is clear that the equilibrium quantities in this model are simply ct = n X i=1 yit . 6. Letting pit denote the price of a share in productive unit i in terms of the consumption good. This is a representative agent economy where the representative consumer has preferences given by E0 ∞ X t=0 β t u(ct ). what prices are depends on the market structure. This limits the ability of consumers to smooth consumption. the output on each productive unit (the dividend) is distributed to the shareholders in proportion to their share holdings. We will suppose an stock market economy. CONSUMPTION AND ASSET PRICING and sometimes consumers cannot borrow on any terms. which drops a random amount of fruit each period. and makes consumption more sensitive to changes in current income. which treats consumption as being exogenous. We can think of each productive unit as a fruit tree. (6. where the representative consumer receives an endowment of 1 share in each productive unit at t = 0. and asset prices as endogenous.78 CHAPTER 6. and the stock of shares remains constant over time. and then shares are traded on competitive markets. and twice differentiable. where yit is the quantity of output produced on productive unit i in period t.

.. and the output vector.t+1 + yi.t+1 !# = 0.18) . pt+1 .17) for i = 1.t+1 + ct = n X i=1 zit (pit + yit ). We have the following envelope conditions: n X ∂v = (pit + yit )u0 [zit (pit + yit ) − pit zi.. the first-order conditions for the optimization problem on the right-hand side of the above Bellman equation are −pit u 0 à n X i=1 [zit (pit + yit ) − pit zi.z t t+1 subject to (6.13). The consumer’s problem is to maximize (6. . . . (6. (6. pt . Lucas (1978) shows that the value function is differentiable and concave. Letting pt . y2t ..14) in the objective function to obtain v(zt .12) subject to (6.15) using (6.. ASSET PRICING 79 quantity of shares in productive unit i held at the beginning of period t.t+1 ! (6. for example yt = (y1t .15) for i = 1.t+1 + yi. zt . . yt+1 )] .t+1 )u ! " 0 à n X i=1 yi. n.2. we can specify a value function for the consumer v(zt . 2. (6. or βu0 (ct+1 ) = 0. and yt denote the price vector. pt . yt+1 ) .16) Substituting in (6. 1.t+1 ] + βEt v(zt+1 .14). and (6..t+1 ] ∂zit i=1 (6..16) then gives −pit u 0 à n X i=1 yit + βEt (pi. ∂zi. and we can substitute using (6. ! ) Now. n.14).. 2. yt ) = cmax [u(ct ) + βEt v(zt+1 . and write the Bellman equation associated with the consumer’s problem as v(zt .t+1 ] + βEt à ! ∂v = 0. yt ) = max u z t+1 ( à n X i=1 [zit (pit + yit ) − pit zi.t+1 ) 0 u (ct ) " # (6.14). pt . 2..14) for t = 0. the representative consumer’s budget constraint is given by n X i=1 pit zi. the vector of share holdings. pt+1 ..6. yt )... ynt ). pit = Et (pi.

u0 (ct ) s=t+1 ⎡ ∞ X ⎤ (6. the current price of a share is equal to the expectation of the product of the future payoff on that share with the intertemporal marginal rate of substitution. Y ) = E(XY ) − E(X)E(Y ). s0 ≥ s ≥ 0). or. but varies over time since consumption is variable. to get β s−t u0 (cs ) ⎦ pit = Et ⎣ yi. mt = βu0 (ct+1 ) . using repeated substitution and the law of iterated expectations (which states that.18). . the representative consumer will pay a high price for an asset which is likely to have high payoffs when aggregate consumption is low. πit = pit and let mt denote the intertemporal marginal rate of substitution. we can write the current share price for any asset as the expected present discounted value of future dividends. Note here that the discount factor is not constant. Et [Et+s xt+s0 ] = Et xt+s0 .s . pi. we can rewrite equation (6. We can also rewrite (6. for any two random variables. That is. using the fact that. cov(X.19) That is. i. CONSUMPTION AND ASSET PRICING That is. X and Y. Perhaps more revealing is to let πit denote the gross rate of return on share i between period t and period t + 1. for a random variable xt . u0 (ct ) Then.80 CHAPTER 6.18) as Et (πit mt ) = 1. covt (πit .t+1 + yi.e. shares with high expected returns are those for which the covariance of the asset’s return with the intertemporal marginal rate of substitution is low.t+1 . where the discount factors are intertemporal marginal rates of substitution. Therefore. mt ) + Et (πit ) Et (mt ) = 1.

random variable. that is u(c) = c. we get pit = u0 ( Pn βAi i=1 yit ) Therefore.2. . 2. n.t+1 !# = Ai . in the aggregate. . the representative consumer will want to consume more today. That is. it must also be true that pt is i. i. therefore the function is unpredictable given information in period t. n. to hold the supply of available assets).d.6. First. Given (6. ASSET PRICING Examples 81 Equation (6. From (6.t+1 − pit 1 = − 1.. if current dividends on assets are high.. Then.t+1 )u0 " Ã n X i=1 yi.21) Equation (6.17) can be used to solve for prices. pit β # (6.i.d.t+1 + yi. the expression inside the expectation operator in (6.17) and (6. This then implies that Et (pi.t+1 + yi. That is. or Et " pi.t+1 + yi.. A second special case is where there is risk neutrality.20) is a function of pi.20). then the marginal utility of consumption is low. i = 1.21) states that the rate of return on each asset is unpredictable given current information.20) for i = 1. the representative consumer must be induced to consume aggregate output (or equivalently.. However.i.t+1 . we then have pit = βEt (pi.17)...t+1 ). and we will show here how this can be done in some special cases. and the current price of the asset is high. if aggregate output (which is equal to aggregate consumption here) is high. 2. where Ai > 0 is a constant.t+1 and yi.e. but will also wish to save by buying more assets so as to smooth consumption. suppose that yt is an i. the current price is the discount value of the expected price plus the dividend for next period. (6. which is sometimes taken in the . each of which is unpredictable given information in period t. and so asset prices must rise.

y2 . 0 < π < 1. 2. Then. s=t+1 or the current price is the expected present discounted value of dividends. suppose we allow the representative agent to borrow and lend. however. there is only one asset. obtaining βy1 p1 = 1−β βy2 p2 = 1−β Note here that p1 > p2 .82 CHAPTER 6. (6. and that yt is i.17). from (6.19) gives pit = Et ∞ X β s−t yi.i. it is straightforward to price a wide variety of assets in this type of model. Alternative Assets and the “Equity Premium Puzzle” Since this is a representative agent model (implying that there can be no trade in equilibrium) and because output and consumption are exogenous. we will suppose that output takes on only two values. yt = y1 . A third example considers the case where u(c) = ln c and n = 1.21) holds only in the case where the representative consumer is risk neutral. CONSUMPTION AND ASSET PRICING Finance literature as an implication of the “efficient markets hypothesis. That is.s . y1 y1 p1 = β π (p1 + y1 ) + (1 − π) (p2 + y2 ) y1 y2 y2 y2 p2 = β π (p1 + y1 ) + (1 − π) (p2 + y2 ) y1 y2 " " # # Since the above two equations are linear in p1 and p2 . there is a risk-free asset which trades on . with Pr[yt = y1 ] = π. Also.” Note here. that (6. that is the price of the asset is high in the state when aggregate output is high. it is straightforward to solve. Let pi denote the price of a share when yt = yi for i = 1. we obtain two equations which solve for p1 and p2 . Also. that is. which is simply a share in aggregate output. with y1 > y2 . For example.d.

1−γ In the data set which Mehra and Prescott examine. setting pi.17). They consider preferences of the form c1−γ − 1 u(c) = .6.e.2. the average rate of return . we will have bt = 0. an equity share which is a claim to aggregate output. u0 (ct ) # (6. but it is more straightforward to simply use equation (6. ASSET PRICING 83 a competitive market at each date. The representative agent’s budget constraint is then n X i=1 pit zi.23) 1 If the representative agent is risk neutral. qt (6. We wish to determine qt . and a one-period risk-free asset as discussed above. and let qt denote the price of a bond in terms of the consumption good in period t. they have no value at the end of period t + 1) and yi.t+1 + ct + qt bt+1 = n X i=1 zit (pit + yit ) + bt In equilibrium. Let bt+1 denote the quantity of risk-free bonds acquired in period t by the representative agent (note that bt+1 can be negative.t+1 = 0 (since these are one-period bonds. and prices need to be such that the bond market clears. there is a zero net supply of bonds. the representative agent can issue bonds).t+1 = 1 to get qt = βEt " u0 (ct+1 ) . and this can be done by re-solving the consumer’s problem.22) The one-period risk-free interest rate is then rt = 1 − 1. This is a one-period risk-free bond which is a promise to pay one unit of consumption in the following period. then qt = β and rt = β − 1. Mehra and Prescott (1985) consider a version of the above model where n = 1 and there are two assets. that is the interest rate is equal to the discount rate. i. which includes annual data on risk-free interest rates and the rate of return implied by aggregate dividends and a stock price index.

is given by Rt = Et à pt+1 + yt+1 − pt . Suppose that output can take on two values. we first need to determine expected returns. the average equity premium is about 6%. pt ! . but we will illustrate their ideas here in a model where consumption does not grow over time.84 CHAPTER 6. i = 1. that is Pr[yt+1 = yj | yt = yi ] = πij . where 0 < ρ < 1. pi .25) Also. denoted Rt . h h i i (6. so (6. where qt = qi and pt = pi when yt = yi . Here. 2. to determine risk premia. Mehra and Prescott show that this equity premium cannot be accounted for by Lucas’s asset pricing model. For the equity share.27) Now. we have −γ −γ −γ p1 y1 = β ρ(p1 + y1 )y1 + (1 − ρ)(p2 + y2 )y2 . That is. with y1 > y2 . suppose that yt follows a two-state Markov process. Further. CONSUMPTION AND ASSET PRICING on equity is approximately 6% higher than the average rate of return on risk-free debt. (6. for i = 1. In any period. and given by rt in (6.24) and (6. 2. We will assume that πii = ρ.27) give us solutions to the four asset prices. 2. t.24) (6.22) implies that y1 q1 = β ρ + (1 − ρ) y2 y2 q2 = β ρ + (1 − ρ) y1 " à " à !γ # !γ # (6. the return on the risk-free asset is certain.25) are two linear equations in the two unknowns p1 and p2 . −γ −γ −γ p2 y2 = β ρ(p2 + y2 )y2 + (1 − ρ)(p1 + y1 )y1 .26) (6. we want to solve for the asset prices qi . for i = 1. The MehraPrescott argument goes as follows.23). Let rt = ri when yt = yi for i = 1. (6.17).24)-(6. From (6. Mehra and Prescott construct a version of Lucas’s model which incorporates consumption growth. 2. y1 and y2 . the expected return. Now.

a higher γ tends to cause an increase in the average risk-free interest rate. rt . y1 . γ. the higher is γ. and y2 so as to replicate the observed properties of aggregate consumption (in terms of serial correlation and variability). then to find parameters β and γ such that e(β.6. . we get p1 + y1 R1 = ρ p1 p2 + y2 R2 = ρ p2 Ã Ã ! ! p2 + y2 + (1 − ρ) p1 p1 + y1 + (1 − ρ) p2 Ã Ã ! ! −1 −1 Now. ρ. y2 ) ∼ . 2 Therefore. y1 . 2 2 Mehra and Prescott’s approach is to set ρ. it helps to highlight the roles played by γ in this model. which is a primary determinant of the expected return on equity. What they find is that γ must be very large. y1 . That is. and = much outside of the range of estimates for this parameter which have been obtained in other empirical work.06. ASSET PRICING 85 Therefore. which is critical in determining the risk-free rate of interest. the average equity premium is e(β. y2 ) = 1 1 (R1 − r1 ) + (R2 − r2 ) . the value of γ captures risk aversion. First. Given the transition probabilities between output states. the higher is γ the larger is the expected return on equity. as the representative agent must be compensated more for bearing risk. Thus. the unconditional (long-run) probability of being in either state is 1 here. letting Ri denote the expected rate of return on the equity share when yt = yi . and the greater is the tendency of the representative consumer to smooth consumption over time. given plausible levels of risk aversion. γ. ρ. the lower is the intertemporal elasticity of substitution. That is.2. The problem in terms of fitting the model is that there is not enough variability in aggregate consumption to produce a large enough risk premium. γ determines the intertemporal elasticity of substitution. Second. To understand these results. what we are interested in is the average equity premium that would be observed in the data produced by this model over a long period of time.

R.” in the Handbook of Economics of Finance.86 CHAPTER 6.” Econometrica 46.3 References Hall.” Journal of Monetary Economics 15. 889-938. R. (Amsterdam. and Prescott.. Barro ed. 1978.) . R. Handbooks in Economics. 145-162. R. N. R.” Journal of Economic Literature 34. London and New York: Elsevier. Mehra. Hall. MA. 1996.” in Modern Business Cycle Theory. “Consumption. 1989. 1426-1445. Cambridge. and Prescott. Financial markets and asset pricing. Lucas. E. 2003. “Stochastic Implications of the Life Cycle-Permanent Income Hypothesis. 42-71. R. Kocherlakota. “The Equity Premium in Retrospect. Harvard University Press. 971-987. “The Equity Premium: A Puzzle. E. Mehra. 21. 1985. “The Equity Premium: It’s Still a Puzzle. CONSUMPTION AND ASSET PRICING 6.” Journal of Political Economy 86. 1978. Volume 1B. “Asset Prices in an Exchange Economy. vol. North Holland.

Further. employment. (1970). similar to the one studied by McCall (1970). Later. Mortensen and Pissarides developed two-sided search models (for a summary see Pissarides 1990) in which workers and firms match in general equilibrium. Search models have these characteristics. Search theory is a useful application of dynamic programming. explain why unemployment fluctuates and how it is correlated with other key macroeconomic aggregates. and wages are endogenous. and evaluate the efficacy of policies affecting the labor market.Chapter 7 Search and Unemployment Unemployment is measured as the number of persons actively seeking work. we need another set of models. Unemployed workers face a distribution of wage offers which is assumed to be fixed. If we want to understand the behavior of the labor market. there must be frictions which imply that it takes time for an agent to transit between unemployment and employment. job search. Clearly. as we want to study equilibria where agents engage in different activities. We will first study a one-sided search model. These are models of “one-sided” search. i. al. which are partial equilibrium in nature. 87 .e. there is no counterpart to this concept in standard representative agent neoclassical growth models. and then look at a two-sided search model where firms and workers match and bargain over wages. These models need heterogeneity. Some early approaches to search and unemployment are in McCall (1970) and Phelps et. and possibly leisure (not in the labor force).

The parameter δ is referred to as the separation rate. The wage offer is accepted if Ve (w) ≥ Vu and declined otherwise. from the government at the beginning of the period. If an agent is employed receiving ¯ wage w (assume that each job requires the input of one unit of labor each period). consumes it. 1+r max [Ve (w). Assume that b < w. which the worker receives. Assume that w ∈ [0. w. each having preferences given by E0 β t ct . b. Let Vu and Ve (w) denote. Note where 0 < β < 1. then her consumption is also w.2) Ve (w) = β [w + δVu + (1 − δ)Ve (w)] In (7. . where r is the discount rate. These values are determined by two Bellman equations: Vu = β b + ½ Z w ¯ 0 1 . which differ according to the wage. or from effort in searching for a job. as we assume that the worker has no opportunities to save. At the end of the period. the set ¯ [0. b. and then receives a wage offer that she may accept or decline. i. From the point of view of an unemployed agent. There are many different jobs in this economy. The integral in (7. Let β = that we assume that there is no disutility from labor effort on the job. and then receives a wage offer from the distribution F (w). respectively.88 CHAPTER 7. the distribution of wage offers she can receive in any period is given by the probability distribution function F (w). An unemployed worker receives an unemployment benefit.1) (7. SEARCH AND UNEMPLOYMENT 7. the value of being unemployed and the value of being employed at wage w. w].1 A One-Sided Search Model ∞ X t=0 Suppose a continuum of agents with unit mass. w] is the support of the distribution. as of the end of the period.e. Vu ] f (w)dw ¾ (7. the unemployed agent receives the unemployment insurance benefit.1) is the expected utility of sampling from the wage distribution. at the beginning of the period. so that at least some job offers ¯ have higher compensation than the unemployment insurance benefit.1). which has associated with it a probability density function f (w). there is a probability δ that an employed worker will become unemployed.

and then either suffers a separation or will continue to work at the wage w next period.1. we obtain Ve (w) = w + δVu . Similarly. 0 f (w)dw. simplifying.3) using (7. Ve (w) is a strictly increasing linear function of w.2).2) can be simplified to obtain rVe (w) = w + δ[Vu − Ve (w)] (7. if we substitute for Vu in equation (7. and subtract Vu from both sides to obtain rVu = b + Z w ¯ 0 max [Ve (w) − Vu . Note that an employed agent will choose to remain employed if she does not experience a separation. because Ve (w) ≥ Vu . there is some w∗ such that Ve (w) ≥ Vu for w ≥ w∗ .4). (7. the employed agent receives the wage. Thus. r+δ (7.5). otherwise she would not have accepted the job in the first place.3) On the right-hand side of (7.6) r Then. r+δ ∙ ¸ . and Ve (w) ≤ Vu for w ≤ w∗ . For (7. r + δ w∗ ∗ Z w ¯ 0 w − w∗ max .5). we have w∗ Vu = . w∗ = b + ¯ 1 Zw (w − w∗ )f (w)dw. The value w∗ is denoted the reservation wage. consumes it.3) is the flow return when unemployed plus the expected net increase in expected utility from the unemployed state.4) We now want to determine what wage offers an agent will accept when unemployed. an unemployed agent will accept any wage offer of w∗ or more. (7. and decline anything else. substitute 1 β = 1+r .5) Therefore. A ONE-SIDED SEARCH MODEL 89 In (7. we get w =b+ or. That is. In search models.6) and for Ve (w) using (7. (7. w.7. so from (7. The reservation wage satisfies Ve (w∗ ) = Vu . From (7.1). a useful simplification of the Bellman equations is obtained as follows. 0] f (w)dw. divide both sides by β.

while the right-hand side is a decreasing and continuous function of w∗ .1 Comparative Statics It is now straightforward to use equation (7. For w∗ = 0. That is. while it is intuitively clear that an unemployed worker would never accept a job offering a wage smaller than the unemployment insurance benefit.1. We depict the determination of the reservation wage in Figure 7. and it is unique. 7. the reservation wage is w1 . he or she would also not accept a wage offer that exceeds b by a small amount. 1 w =b+ r+δ ∗ ½Z w ¯ w∗ wf (w)dw − w [1 − F (w )] . hoping to receive a wage offer that is much higher in the future. This is because an unemployed worker is willing to turn down such an offer and continue to collect b.7) solves for the reservation wage w∗ . where ¯ 1 Zw [1 − F (w)]dw. SEARCH AND UNEMPLOYMENT and simplifying further.1. and for w = w the left-hand side exceeds the right. the right-hand side of the equation exceeds the left-hand side. w =b+ r + δ w∗ ∗ (7.7) to determine how changes in agents’ preferences and in the environment affect the reservation .90 CHAPTER 7. w =b+ ∗ r+δ w ∗ ½ and simplify again to get ¯ 1 Zw [1 − F (w)]dw. a solution for w exists. ¾ ∗ ∗ ¾ Next integrate by parts to obtain Z w ¯ 1 ∗ ∗ w − w F (w ) − ¯ F (w)dw − w∗ [1 − F (w∗ )] . A(w ) = r + δ w∗ ∗ ∗ In the figure.7) Equation (7. ¯ ∗ Therefore. Note from the figure that ∗ we must have w1 > b. Note that the left-hand side of this equation is a strictly increasing and continuous function of w∗ .

1: Determination of the Reservation Wage w* b+A(0) b+A(w1*) b+A(w*) b 0 w1* _ w reservation wage w* .Figure 7.

is that unemployed workers become less picky about the jobs they will accept. What equation (7. An unemployed worker therefore becomes more picky concerning the jobs he or she will accept. If the separation rate δ increases.5) and (7. . consider a change in the unemployment insurance benefit. r+δ This effect occurs because higher δ implies that the expected lifetime of a job is lower. because it is not so tempting to hold out for a better job that will now tend to dissolve more quickly.7) and solving gives dw∗ r+δ = > 0. the reservation wage increases with an increase in the unemployment insurance benefit. They become less picky and reduce their reservation wage. Next.1. totally differentiating (7. Totally differentiating equation (7. then agents discount future payoffs at a higher rate. as shown in Figure 7. note from equation (7. this will reduce the difference between the value of being employed and the value of being unemployed (given the reservation wage). db r + δ + 1 − F (w∗ ) Therefore. Other experiments that we could consider involve changes in the distribution of wage offers F (w).7.6) is Ve (w) − Vu = w − w∗ . so we can kill two birds with one stone. perhaps counterintuitively.7) that r and δ will affect the determination of the reservation wage in exactly the same way. and therefore are less willing to wait for a better wage offer in the future. This occurs because an increase in b reduces the cost of search while unemployed. dr dδ (r + δ) [r + δ + 1 − F (w∗ )] w∗ Therefore.2. in increase in either r or δ reduces the reservation wage. A ONE-SIDED SEARCH MODEL 91 wage w∗ .7) tells us is that the wage offer distribution matters for the determination of the reservation wage w∗ in terms of how it affects G(F ) = Z w ¯ w∗ [1 − F (w)]dw. The effect of all jobs being less attractive. b. First. which from (7. If r increases.7) in a similar fashion to what we did for a change in b to get Z w ¯ dw∗ dw∗ −1 = = [1 − F (w)]dw < 0.

2: An increase in b increases the reservation wage w* b1+A(0) b1+A(w1*) b2+A(w*) b1+A(w*) b 0 w1* w2* reservation wage w* .Figure 7.

92

CHAPTER 7. SEARCH AND UNEMPLOYMENT

That is, if a change in F increases G(F ), then it has a qualitative effect on the reservation wage identical to the effect of an increase in b, as in Figure 7.2. That is, w∗ increases. This would be the effect if, for example, there were a first-order-stochastic-dominance shift in F (w), whereby F (w) decreases for all w ∈ (0, w). Thus, if the wage ¯ distribution improves in the sense of first-order stochastic dominance, then w∗ must increase because the expected gain from turning down a wage offer and waiting for a better one increases. Note also that G(F ) can increase if the dispersion in the distribution F (w) increases in particular ways. For example, if dispersion increases in such a way that the probability mass to the right of the initial w∗ remains the same (i.e. F (w∗ ) does not change for the initial w∗ ), then G(F ) increases and w∗ must increase.

7.1.2

Employment and Unemployment

Now that we have determined the behavior of individual agents, as summarized by how w∗ is determined, we can say something about the behavior of aggregate employment and unemployment. Now, let ut denote the fraction of agents who are unemployed in period t. The flow of agents into employment is just the fraction of unemployed agents multiplied by the probability that an individual agent transits from unemployment to employment, ut [1 − F (w∗ )] . Further, the flow of agents out of employment to unemployment is the number of separations (1 − ut )δ. Therefore, the law of motion for γt is ut+1 = ut − ut [1 − F (w∗ )] + (1 − ut )δ = ut [F (w∗ ) − δ] + δ. (7.8)

Since | F (w∗ ) − δ |< 1 , ut converges to a constant, u, which is determined by setting ut+1 = ut = u in (7.8) and solving to get u= δ . δ + 1 − F (w∗ ) (7.9)

Therefore, the number of unemployed increases as the separation rate increases, and as the reservation wage increases (though note that the reservation wage also depends on the separation rate). That is, a higher

7.1. A ONE-SIDED SEARCH MODEL

93

separation rate increases the flow from employment to unemployment, increasing the unemployment rate, and a higher reservation wage reduces the job-finding rate, 1 − F (w∗ ), thus reducing the flow from unemployment to employment and increasing the unemployment rate. We can conclude, from our analysis of what affects the reservation wage w∗ , that an increase in b or a decrease in r, which each increases the reservation wage, will also increase the unemployment rate, from (7.9). An increase in the separation rate δ has the direct effect of increasing the unemployment rate, but it also will reduce the reservation wage, which will reduce the unemployment rate. The net effect is ambiguous. Similarly, a first-order stochastic dominance shift in the wage offer distribution F (w) has the effect of increasing the reservation wage and therefore reducing the unemployment rate, but since F (w) falls for each w ∈ (0, w), the net effect on F (w∗ ) is ambiguous. The unem¯ ployment rate could increase or decrease. However, if F (w) changes in such a way that dispersion increases while holding F (w∗ ) constant for the initial w∗ , then w∗ increases, F (w∗ ) increases, and the increase in dispersion increases the unemployment rate.

7.1.3

An Example

Suppose that there are only two possible wage offers. An unemployed agent receives a wage offer of w with probability π and an offer of zero ¯ with probability 1 − π, where 0 < π < 1. Suppose first that 0 < b < w. ¯ Here, in contrast to the general case above, the agent knows that when she receives the high wage offer, there is no potentially higher offer that she foregoes by accepting, so high wage offers are always accepted. Low wage offers are not accepted because collecting unemployment benefits is always preferable, and the agent cannot search on the job. Letting Ve denote the value of employment at wage w, the Bellman equations ¯ can then be written as rVu = b + π(Ve − Vu ), rVe = w + δ(Vu − Ve ), ¯

94

CHAPTER 7. SEARCH AND UNEMPLOYMENT

and we can solve these two equations in the unknowns Ve and Vu to obtain (r + π)w + δb ¯ , Ve = r(r + δ + π) Vu = Note that π w + (r + δ)b ¯ . r(r + δ + π)

w−b ¯ r+δ+π depends critically on the difference between w and b, and on the dis¯ count rate, r. The number of unemployed agents in the steady state is given by δ u= , δ+π so that the number unemployed decreases as π increases, and rises as δ increases. Now for any b > w, clearly we will have γ = 0, as no offers of ¯ employment will be accepted, due to the fact that collecting unemployment insurance dominates all alternatives. However, if b = w, then an ¯ unemployed agent will be indifferent between accepting and declining a high wage offer. It will then be optimal for her to follow a mixed strategy, whereby she accepts a high wage offer with probability η. Then, the number of employed agents in the steady state is Ve − Vu = u= δ , δ + ηπ

which is decreasing in η. This is a rather stark example where changes in the UI benefit have no effect before some threshold level, but increasing benefits above this level causes everyone to turn down all job offers.

7.1.4

Discussion

The partial equilibrium approach above has neglected some important factors, in particular the fact that, if job vacancies are posted by firms, then the wage offer distribution will be endogenous - it is affected by the rate at which the unemployed accept which jobs, and by what

7.2. A TWO-SIDED SEARCH MODEL

95

types of jobs are posted by firms. In addition, we did not take account of the fact that the government must somehow finance the payment of unemployment insurance benefits. A simple financing scheme in general equilibrium would be to have UI benefits funded from lump-sum taxes on employed agents.

7.2

A Two-Sided Search Model

For many macroeconomic issues, we want general equilibrium search models of unemployment in which we can determine wages endogenously and seriously address the effects of policy. Versions of two-sided search and matching models, developed first in the late 1970s, have been used extensively in labor economics and macro. For further references see Mortensen (1985), Pissarides (1990), and Rogerson, Shimer, and Wright (2005). What I have done here borrows heavily from the latter survey, though I work here exclusively in discrete time.

7.2.1

The Model

There is a continuum of workers with unit mass, each of whom has preferences given by ∞ X µ 1 ¶t ct , E0 t=0 1 + r where ct is consumption and r > 0. There is also an infinite mass of firms, with each firm having preferences given by E0
∞ Xµ t=0

1 1+r

¶t

(πt − xt ),

where πt denotes the firm’s profits, which are consumed by the firm, and xt denotes any disutility from posting a vacancy during period t. Goods are perishable, and savings is assumed to be zero for each agent in each period. Let ut denote the mass of workers who are unemployed each period, with 1 − ut being the mass of workers who are matched with firms, producing output, and therefore employed. As well, vt is the mass of

96

CHAPTER 7. SEARCH AND UNEMPLOYMENT

firms which post vacancies in period t. Each period, there are matches between unemployed workers and firms posting vacancies, with mt denoting the mass of matches according to mt = m(ut , vt ), where m(·, ·) is the matching function. Assume that m(·, ·) is continuous, increasing in both arguments, concave, homogeneous of degree 1, and that m(0, v) = m(u, 0) = 0 for all v, u ≥ 0. The probability with which an individual unemployed worker is matched with a firm v posting a vacancy in period t is given by m(utt,vt ) = m(1, ut ) given that u t the matching function satisfies homogeneity of degree 1. Similarly, the probability that an individual firm posting a vacancy is matched with v t a worker is m(utt,vt ) = m( ut , 1). For convenience, define θt ≡ ut , where θt v v t is a measure of labor market tightness in period t, in that an increase in θt increases the job-finding probability for an unemployed worker, and lowers the probability that a firm can fill a job. Assume that 1 lim m , 1 = lim m(1, θ) = 1. θ→0 θ→∞ θ Each firm has a technology for producing output. With this technology, y units of output can be produced with one unit of labor input each period, and zero units of output for any other quantity of labor input. Each worker has one unit of time available each period. When a worker and firm meet, and agree to a contract, they can then jointly produce y units of output until they become separated. Separation occurs each period with probability δ. While unemployed, a worker receives unemployment insurance compensation of b each period (note that, as in the one-sided model, we don’t account for the financing of b by the government). A firm posting a vacancy incurs a cost in terms of utility of k each period the vacancy is posted. Any firm not posting a vacancy and not matched with a worker receives zero utility.
µ ¶

7.2.2

Bargaining

We will confine attention to steady state equilibria where ut = u and vt = v for all t. When a worker and firm meet, they will negotiate a

Here. A tractable approach to the determination of the equilibrium wage is to suppose that the firm and worker engage in Nash bargaining. is J(y − w) = 1 [y − w + (1 − δ)J(y − w) + δV ] . let U denote the value to the worker of remaining unemployed. with 0 ≤ α ≤ 1. Note that the above optimization problem is not a problem solved by any individual agent . A TWO-SIDED SEARCH MODEL 97 wage w. and J(y − w) − V denotes the surplus from the match for the firm. The worker and the firm can only come to an agreement if W (w) − U ≥ 0 and J(y − w) − V ≥ 0 for some w. as of the end of the period. The total surplus is the sum of these two quantities. where α is a parameter which is a measure of the worker’s bargaining power. so that w = arg max [W (w0 ) − U] [J(y − w0 ) − V ]1−α 0 w α subject to J(y − w0 ) − V ≥ 0. Ignoring the constraints in the above optimization problem for now. 1+r . the first-order condition for a maximum simplifies to give αW 0 (w)[J(y − w) − V ] − (1 − α)J 0 (y − w)[W (w) − U] = 0. which is the payment that will be made to the worker in each period until the firm and worker are separated. where W (w) − U denotes the surplus from the match for the worker. all values are defined to be as of the end of the period.7. and V the value to the firm of posting a vacancy.instead the solution to this problem describes the outcome of bargaining between the worker and the firm. W (w0 ) − U ≥ 0. 1+r and the value of a match for a firm. the value of being employed at wage w. and let J(y − w) denote the value of the match to the firm. (7. is given by W (w) = 1 [w + (1 − δ)W (w) + δU] . or W (w) + J(y − w) − U − V.10) For a worker. Let W (w) denote the value of the match to a worker if the wage is w. As well.2. given the wage w.

Since in equilibrium all jobs will pay the same wage. 1 1+r θ θ V . but this is independent of the wage that is being negotiated in the particular labor contract between an individual worker and an individual firm.11) and (7. Then. gives. 1+r ½ µ ¶ ∙ µ ¶¸ 1 1 1 −k + m . we need Bellman equations determining values for an unemployed worker and for a firm posting a vacancy. ¾ . Therefore. (7.11) 1 and so W 0 (w) = J 0 (y−w) = r+δ .12) (7.3 Equilibrium Next. r+δ y − w + δV .12).10) simplifies to α[J(y − w) − V ] − (1 − α)[W (w) − U ] = 0. respectively. by U= and V = 1 {b + m(1.98 CHAPTER 7. equation (7. SEARCH AND UNEMPLOYMENT Simplifying these two Bellman equations. we obtain. W (w) = and J(y − w) = w + δU . θ)]U } . Further. Note here that U and V will in general depend on the wages paid by other firms.2. θ)W + [1 − m(1. we will let W denote the equilibrium value of being employed for a worker and J the value of a match for a firm. respectively. just as we did for the onesided search model. respectively. suppose that any meeting between a firm and worker results in a successful match.13) 7. r+δ (7. U and V are determined. from (7. rW (w) = w + δ[U − W (w)] and rJ(y − w) = y − w + δ[V − J(y − w)]. 1 J + 1 − m . Therefore.

determined by the worker’s bargaining power. (7.21) and from (7. subtracting (7.14) from (7. θ)α (7.18) that is Nash bargaining implies here that the worker gets a constant fraction α of the total surplus.15). That is. (7. it follows that J − V = (1 − α)S.1 θ ´. and given V = 0. A TWO-SIDED SEARCH MODEL or simplifying. firms have to be indifferent between their alternative opportunity.14) µ ¶ 1 rV = −k + m . r(W + J − U − V ) = y − b − k + δ(U − W − J) − m(1.20) Then from (7. (7. (7. and posting a vacancy. where S =W +J −U −V =W +J −U (7. That is V = 0. r + δ + m(1.13) gives W − U = αS.19). we get S= k (1 − α)m ³ 1 . which yields zero value. θ)(W − U) (7. (7. (7.12).22) .20) to get S= y−b . (7. in a steady state equilibrium.15) θ The final detail we need in the model is the analog of a zero-profit condition for firms. Therefore.14) imply. and (7.7.17) Then. and (7.12). (7.17).19) Next.11).11) plus (7.19). 1 (J − V ).16) Let S denote the total surplus from a match for a firm and a worker. 99 rU = b + m(1.18). (7.2. θ)(W − U). (7. we can simplify (7. equation (7.

23) w + δαS . then the equilibrium is unique. F (∞) = r+δ+α . From (7. which in turn implies that both a matched worker and a matched firm earn positive surplus. (7. where S = S ∗ and θ = θ∗ in equilibrium. r+δ and G(∞) = ∞. given θ. Then. r (7. θ) + δ and given the definition of θ. let F (θ) denote the right-hand side of (7.26) m(1. (7.12). θ) + δ U= Now.21) and (7. while the flow of workers from employment to unemployment is (1 − u)δ. which implies that. our conjecture that each meeting between a firm and a worker results in a successful match is correct. Therefore.22) solve for S and θ. then Equations (7. as in Figure 7. (7. (7.22). if and only if the cost of posting a vacancy is sufficiently small. r+δ+α r+δ so S > 0 in equilibrium. If this condition holds. The functions F (·) and G(·) are continuous y−b k with F 0 (θ) < 0 and G0 (θ) > 0. SEARCH AND UNEMPLOYMENT then given S and w. F (0) = y−b . these flows are equal. In a steady state. u is given by δ u= . we then have δθ . . θ).27) v = uθ = m(1. an equilibrium exists if and only if k< (1 − α)(y − b) . the flow of workers from unemployment to employment is um(1. r+δ that is. G(0) = 1−α . and we will have y−b y−b < S∗ < . Therefore.11) gives W = and since W − U = αS.25) r In the steady state.21) and G(θ) the right-hand side of (7. (7. given S the wage is determined by w = y − (r + δ)(1 − α)S.3.100 CHAPTER 7. we can solve for all other endogenous variables.24) w + (δ − r)αS .

θ)α) (0.0) θ∗ θ .1)] S (y-b)/(r+δ) S* (y-b)/(r+δ+α) k/[1-α] (y-b)/(r+δ+m(1.3: Determination of S and θ k/[(1-α)m(1/θ.Figure 7.

and the unemployment rate falls. interpreted as an increase in aggregate productivity. The decrease in θ causes v to fall. This increases the job-finding rate for unemployed workers. 2 dθ [m(1. differentiate (7.2. In Figure 7.3. An increase in unemployment insurance compensation acts to reduce total surplus in a match and therefore makes posting vacancies less attractive for firms. an increase in y will result in an increase in S and an increase in θ. θ) + δ = = > 0. this has the effect of reducing both S and θ. A decrease in b has the opposite effects of an increase in y. making posting vacancies more attractive for firms. From Figure 7. referred to as a Beveridge curve. θ) + δ − m2 (1. unemployment must then fall. Finally.27) with respect to θ to get dv m(1.23) that the wage rises. so that v and θ increase. Though we are looking at a steady state equilibrium. θ) + δ]2 which uses the homogeneity-of-degree-one property of the matching function.2.7.4 Experiments Consider first a change in y. From (7. and will tend to yield a negative correlation between u and v.26). which is not observed in the . The increase in productivity makes unemployment and vacancies move in opposite directions. so that v and θ fall.3. It can also be shown that the wage w increases. but the direct effect of δ on v is for v to rise. As for the case of changes in productivity. since unemployment is more attractive for workers. so that shocks to the separation rate may tend to produce a positive correlation between u and v. The mechanism at work here is that an increase in y will tend to increase the total surplus from a match. consider an increase in the separation rate δ. the mechanism at work here also transfers to stochastic environments. Unemployment u must rise. Note from equation (7. and firms therefore have to pay higher wages to make employment sufficiently attractive for workers. and because of the decrease in θ which reduces the job-finding rate. To determine the effect on vacancies. θ) + δ] [m(1. θ)θ m1 (1. A TWO-SIDED SEARCH MODEL 101 7. this mechanism works similarly in stochastic versions of two-sided search models. This reduces the job-finding rate and u rises. both because of the direct effect of δ on u. and it is possible for u and v to both rise.

SEARCH AND UNEMPLOYMENT data. . 113-126. so this is not quite a general equilibrium model.” Quarterly Journal of Economics 84. Journal of Monetary Economics 36. workers and firms have difficulty matching in practice because there is heterogeneity on both sides of the market. this will in general cause firms and workers to match at a different rate. we also need to be more serious about savings. These applications include Andolfatto (1995). at least in terms of qualitative features of the data. 269-300. A fundamental weakness in the standard two-sided matching model is the matching function specification. productivity shocks do a better job than do separation rate shocks. For example.2. M. 1995. Therefore. we have left out the details of the financing of unemployment insurance payments. 1970.3 References Andolfatto. Merz (1995). “Search in the Labor Market and the Real Business Cycle”. “Economics of Information and Job Search. That is. McCall. and capital accumulation. That is. investment. J.102 CHAPTER 7. and Shimer (2005). For this model to successfully address problems in business cycle behavior and policy (such as the optimal design of unemployment insurance systems).5 Discussion As with the previous one-sided search model.” American Economic Review 86. and because there is private information about worker types and firm types. “Business Cycles and Labor Market Search. Merz. 7. There have been a number of interesting applications of stochastic two-sided search models to business cycle problems. This is basically a cheap way to capture heterogeneity in the model without specifying it explicitly. The matching function is not likely to be immune from the Lucas critique in many policy applications. 112-132. 7. if government labor market policy changes. We would not expect the function to be invariant to changes in policies. the matching function is not a structural object. 1995. D.

Norton. REFERENCES 103 Phelps. “The Cyclical Behavior of Equilibrium Unemployment and Vacancies. 2005. 1990.. al. University of Chicago. Shimer.” manuscript. . Equilibrium Unemployment Theory. Cambridge. MA. E. Shimer. et. New York. Pissarides. R. and Wright. R. D. 25-49. Amsterdam. 2005. “Job Search and Labor Market Analysis. 1970. Microeconomic Foundations of Employment and Inflation Theory.7. R. “Search-Theoretic Models of the Labor Market: A Survey.. Rogerson.3. North-Holland. and University of Pennsylvania. C.” American Economic Review 95. Mortensen.” in Handbook of Labor Economics. 1985. Arizona State University. R. Basil Blackwell.

SEARCH AND UNEMPLOYMENT .104 CHAPTER 7.

and has been widely-used. This cash-in-advance approach was pioneered by Lucas (1980. which we will study in this chapter. 1982). strictly concave. this is a static representative agent model. and ns is labor supply.g. Assume t that u(·) is strictly increasing. particularly in quantitative work (e. and twice differentiable. 8. The representative consumer has preferences given by E0 ∞ X t β [u(c t=0 t) − v(ns )] .Chapter 8 A Cash-In-Advance Model Many macroeconomic approaches to modeling monetary economies proceed at a higher level than in monetary models with search or overlapping generations (one might disparagingly refer this higher level of monetary model as implicit theorizing).1 A Simple Cash-in-Advance Model With Production In its basic structure. is to simply assume that money accumulated in the previous period is necessary to finance current period transactions. Another approach. One approach is to simply assume that money directly enters preferences (money-in-the-utility-function models) or the technology (“transactions cost” models). 105 .1) where 0 < β < 1. Cooley and Hansen 1989). ct is consumption. t (8. with an added cash-in-advance constraint which can potentially generate dynamics.

2) where yt is output. The government budget constraint takes the form ¯ ¯ Mt+1 = Mt + Pt τt . . The representative firm has a constant-returns-to-scale technology. a real bond issued in period t is a promise to pay one unit of the consumption good when the asset market opens in period t + 1. nd is labor input.3) ¯ where Mt is the money supply in period t.106 CHAPTER 8. and zt one-period real bonds. The asset market opens. on which the consumer can exchange money. 4. and γt is a random technology t shock. 2. Money enters the economy through lump-sum transfers made to the representative agent by the government. Assume that ¯ ¯ (8.4) Mt+1 = θt Mt . Bt one-period nominal bonds. A CASH-IN-ADVANCE MODEL and that v (·) is increasing. and receives a cash transfer from the government. The asset market closes and the consumer supplies labor to the firm. t (8. where θt is a random variable. Here. where h is the endowment of time the consumer receives each period. and real bonds. nominal bonds. Similarly. The consumer enters the period with Mt units of currency. Each nominal bond issued in period t is a promise to pay one unit of currency when the asset market opens in period t + 1. with v 0 (0) = 0 and v 0 (h) = ∞. 3. the timing of transactions can be critical to the results. the timing of events within a period is as follows: 1. Pt is the price level (the price of the consumption good in terms of money) and τt is the lump-sum transfer that the representative agent receives in terms of consumption goods. strictly convex. (8. the current period shocks. In cash-in-advance models. yt = γt nd . and twice differentiable. The consumer learns θt and γt .

1. γt+1 )] t t (8. bt .8) t Note that we have used (8. θt+1 . The goods market closes and consumers receive their labor earnings from the firm in cash. zt . The first is a “cash-in-advance constraint. and qt is the price in units of the consumption good of a newly-issued real bond. St Bt+1 + Pt qt zt+1 + Pt ct ≤ Mt + Bt + Pt zt + Pt τt . we will assume throughout that (8.bt+1 . and later establish conditions that will guarantee this. St Bt+1 + Pt qt zt+1 + Mt+1 + Pt ct ≤ Mt + Bt + Pt zt + Pt τt + Pt wt ns .6) can then be rewritten as ¯ St bt+1 θt + pt qt zt+1 + pt ct ≤ mt + bt + pt zt + pt τt and St bt+1 θt + pt qt zt+1 + mt+1 θt + pt ct ≤ mt + bt + pt zt + pt τt + pt wt ns (8.8).e.ns .7) may not bind.5) and (8. The second constraint is the consumer’s budget constraint. To make the consumer’s dynamic optimization stationary.6) t where wt is the real wage. γt ) = maxct . it is useful to divide through constraints (8. Constraints (8.8) will be binding at the optimum. the nominal money supply.4) to simplify (8. γt ).1) subject to two constraints. However. The consumer’s problem is to maximize (8.5) Here. A SIMPLE CASH-IN-ADVANCE MODEL WITH PRODUCTION107 5.8. 6. The constraint (8.7) . The goods market opens.5) and (8.mt+1 . The Bellman equation is then v(mt . bt+1 .7) and (8. and to change variables. θt . The consumer’s optimization problem can be formulated as a dynamic programming problem with the value function v(mt . zt . zt+1 . where consumers purchase consumption goods with cash. the constraint that the consumer must finance consumption purchases and purchases of bonds from the asset stocks that she starts the period with. bt . (8.6) ¯ by Mt .” i. but (8.zt+1 [u(ct ) − v(ns ) + βEt v(mt+1 . St is the price in units of currency of newly-issued nominal bond. for example P pt ≡ Mtt . defining lower-case variables (except for previously-defined real variables) to be nominal variables scaled by the nominal money supply. θt .7) binds. (8.

we have λt = μt (1 − St ). bt+1 .12) (8.108 CHAPTER 8.15) (8. is less than one. and (8.14). (8. Assuming that the value function is differentiable and concave. zt+1 . t ∂ns t ∂v ∂L = βEt − μt θt = 0.9) ∂ct ∂L = −v 0 (ns ) + μt pt wt = 0.7) and (8. Therefore. St − 1 > 0. ∂mt ∂bt ∂v = (λt + μt ) pt . ∂L = u0 (ct ) − λt pt − μt pt = 0. St . (8.14) (8. .10) (8. From (8. θt+1 .12). the cash-in-advance constraint binds if and only if the price of the nominal bond. The Lagrangian for the optimization problem on the right-hand side of the Bellman equation is L =u(ct ) − v(ns ) + βEt v(mt+1 . ∂mt+1 ∂mt+1 ∂L ∂v = βEt − λt St θt − μt St θt = 0. This implies that the nominal 1 interest rate. γt+1 ) t +λt (mt + bt + pt zt + pt τt − St bt+1 θt − pt qt zt+1 − pt ct ) +μt (mt + bt + pt zt + pt τt + pt wt ns − St bt+1 θt − pt qt zt+1 − mt+1 θt − pt ct ).11).8). t where λt and μt are Lagrange multipliers.13) A binding cash-in-advance constraint implies that λt > 0. ∂zt (8. ∂bt+1 ∂bt+1 ∂v ∂L = βEt − λt pt qt − μt pt qt = 0. the unique solution to this optimization problem is characterized by the following first-order conditions. A CASH-IN-ADVANCE MODEL subject to (8. ∂zt+1 ∂zt+1 We have the following envelope conditions: ∂v ∂v = = λt + μt .11) (8.

u0 (ct+1 ) Pt+1 ! = St u0 (ct ) . pt+1 pt wt ! ! (8. pt+1 pt Given the definition of pt . i.17) more informatively as à ! u0 (ct+1 )Pt wt = v 0 (ns ) (8. Pt (8.e.16) and (8. ns = nd = nt . (8. use (8. In equation (8. and the left-hand side is the discounted expected marginal utility of labor earnings. this period’s labor earnings cannot be spent until the following period. Profit maximization by the representative firm implies that wt = γt in equilibrium.16) (8. play no role in determining the equilibrium.20) is a pricing equation for the nominal bond. in equilibrium the labor market clears. we can write (8.18) is a familiar pricing equation for a risk free real bond. The right-hand side is the marginal cost.22) (8.15) to substitute for the partial derivatives of the value function in (8.e. t t ¯ the money market clears. i.21) .20). (8.10) to substitute for the Lagrange multipliers to obtain βEt βEt à u0 (ct+1 ) v0 (ns ) t − θt = 0. Note that the asset pricing relationships.1. the right-hand side is the marginal disutility of labor.8.14) and (8. Also.18) and (8. Equation (8.11)-(8.13).19) βEt t Pt+1 and βEt à βEt u0 (ct+1 ) − qt u0 (ct ) = 0.18) à u0 (ct+1 ) u0 (ct ) − St θt = 0. and the left-hand side is the expected utility of the payoff on the bond in period t + 1.23) (8.17) (8. equation (8.20) Now. in terms of foregone consumption.19).9) and (8. A SIMPLE CASH-IN-ADVANCE MODEL WITH PRODUCTION109 Now. from purchasing a nominal bond in period t. Mt = Mt or mt = 1. and then use (8.

θt Mt is equal to 1.27) is the stochastic law of motion for employment in equilibrium.26). (8. we can then work backward. θt+1 (8.25) and (8.8) (with equality).26). substituting for ct and pt in (8.28) and consumption from (8.16) using (8. and there are regularities in the behavior of velocity over the business cycle which we would like our models to explain. and (8. Empirically. (8. Note that (8. from (8.28) implies that the income velocity of money. (8.3). γt nt (8.21)-(8.3). we also have ct = γt nt . A CASH-IN-ADVANCE MODEL and the bond markets clear. using (8. In this and other cash-in-advance models.21)-(8.26) Now. (8. This equation can be used to solve for nt as a function of the state (γt .4). the velocity of money is a measure of the intensity with which the stock of money is used in exchange. (8.24) Given the equilibrium conditions (8. or.27) Here. defined by Vt ≡ Pt yt ¯ .25). Pt = ¯ θt Mt . θt ).24). (8. we get βEt γt+1 nt+1 u0 (γt+1 nt+1 ) − nt v 0 (nt ) = 0.25). Once nt is determined.24). pt γt nt = θt .4). " # (8. to obtain the price level. (8. and (8. the velocity of money is fixed if the cash-in-advance constraint .25) (8. Also.7) (with equality) give pt ct = θt . bt = zt = 0.110 CHAPTER 8.

γt u0 (γt nt ) (8. and the money supply grows at a constant rate. n is the solution to βγu0 (γn) (8. # (8. i. from (8.34) .17) and (8. that is the money growth factor must be greater than the discount factor. we can also obtain a simple expression for the price of the nominal bond.18) using (8.31) Note that. Substituting for pt and ct in the asset pricing relationships (8. Then.4). there are no technology shocks.e.30) From (8. for the cash-in-advance constraint to bind.25) and (8.2. for our maintained assumption of a binding cash-in-advance constraint to be correct. θt+1 βEt u0 (γt+1 nt+1 ) − qt u0 (γt nt ) = 0.33) − v0 (n) = 0.28) and (8. note that. as the stock of money turns over once per period. where. where γ and θ are positive constants.28) and (8. from (8. St = v0 (nt ) . the price level is given by Pt = ¯ θt+1 M0 . θ Now.8. nt = n for all t. From (8.1 Examples Certainty Suppose that γt = γ and θt = θ for all t. γn (8. This can be viewed as a defect of this model.27).2.2 8.29). (8. EXAMPLES 111 binds.29) (8. or that the equilibrium solution satisfy v0 (nt ) < γt u0 (γt nt ).26) gives βEt " γt+1 nt+1 u0 (γt+1 nt+1 ) − St γt nt u0 (γt nt ) = 0.32) 8. we require that St < 1.32) we must have θ > β.

Comparative statics in equation (8. However. has no effect on any real variables.35). an increase in the money growth rate implies a one-for-one increase in the inflation rate.29) and (8. then this does have real effects. i. A CASH-IN-ADVANCE MODEL and the inflation rate is πt = Pt+1 − 1 = θ − 1.33). With a higher inflation rate. and consumption decrease with the increase in the money growth rate through a labor supply effect. and he/she substitutes leisure for labor.34). from (8. i. Labor earnings are paid in cash. there is a level effect on the price level of a change in θ. money is not super-neutral in this model. money is neutral in the sense that changing the level of the money ¯ supply.e. dθ θβγ 2 u00 (γn) − θ2 v 00 (n) Note also that.34). changing M0 . from (8.35) Here. due to the change in n. if θ increases. the representative agent’s real wage falls. Employment. From (8. Pt (8.112 CHAPTER 8. qt β β θ . but only ¯ increases all prices in proportion (see 8. an increase in the money growth rate causes an increase in the inflation rate. and a direct growth rate effect through the change in θ. which cannot be spent until the following period. That is. Note that M0 does not enter into the determination of n (which determines output and consumption) in (8. With regard to asset prices. and in the intervening time purchasing power is eroded. output.30) we get qt = β and St = The real interest rate is given by rt = 1 1 − 1 = − 1. if the monetary authority changes the rate of growth of the money supply.33) gives dn βγu0 (γn) = < 0. which effectively acts like a tax on labor earnings.e.

i.36) which is a good approximation if β is close to 1. there exists a competitive equilibrium where nt is also i. random variables.i. we have Rt − rt = θ−1 ∼ = θ − 1 = πt . Here. employment. consumption. the current money growth rate provides no information about future money growth. (8. Then. or real and nominal interest rates. and the nominal interest rate is 1 θ Rt = −1= −1 St β Therefore.37) βψ − nt v 0 (nt ) = 0. as this well in general affect ψ and ω.. 8. and (8.i.29) gives (8. given that θt is i. that is the difference between the nominal interest rate and the real interest rate is approximately equal to the inflation rate. β (8.36) is a Fisher relationship. where ψ is a constant. In this model.. Then. where n is a constant. (8.37) implies that nt = n.39) that θt has no effect on output. the real interest rate is equal to the discount rate. (8. suppose that θt and γt are each i. with no effect on the real rate.e.39) where ω is a constant. Increases in the inflation rate caused by increases in money growth are reflected in an approximately one-for-one increase in the nominal interest rate.d. Note in (8.8.2 Uncertainty Now.d. From (8.2. Note however that the probability distribution for θt is important in determining the equilibrium.2. (8.38) .29) and (8. monetary policy has no effect except to the extent that it is anticipated. and so there are no real effects. Here.37)-(8. EXAMPLES 113 i.d. we obtain βψ − St γt nu0 (γt n) = 0 and βω − qt u0 (γt n) = 0.30).

From (8. high γt implies high output and consumption. if the coefficient of relative risk aversion is greater than one. That is. 8. Which effect dominates depends on the strength of the consumption-smoothing effect. Suppose that the monetary authority chooses an optimal money growth policy ∗ θt so as to maximize the welfare of the representative consumer. The social planner solves max ∞ ∞ X t=0 {nt }t=0 β t [u(γnt ) − v(nt )] . . for all t. which increases as curvature in the utility function increases.38). as inflation is expected to be higher. Second. otherwise the nominal interest rate rises. Since yt = γt n.3 Optimality In this section we let γt = γ. Pt . from (8. From (8. To do so. a constant. and this pushes up the price of nominal bonds. consumers buy nominal bonds in order to consume more in the future as well as today. St rises (the nominal interest rate falls). will have real effects here.28). reducing the nominal interest rate.114 CHAPTER 8. First. there is a positive anticipated inflation effect on the nominal interest rate. We want to determine the properties of this optimal growth rule. and allow θt to be determined at the discretion of the monetary authority. Comparative statics gives St γt nu00 (γt n) dSt +1 =− dγt γt u0 (γt n) " # Therefore. the nominal interest rate will tend to fall due to the same forces that cause the real interest rate to fall. A CASH-IN-ADVANCE MODEL The technology shock. so that consumers expect higher inflation. γt . There are two effects on the nominal interest rate. the increase in output causes a decrease in the price level. first consider the social planner’s problem in the absence of monetary arrangements. and qt rises (the real interest rate falls) as the representative consumer attempts to smooth consumption into the future. the increase in output results in a decrease in the marginal utility of consumption.39). The effect on the nominal interest rate. is ambiguous.

4 Problems With the Cash-in-Advance Model While this model gives some insight into the relationship between money.” Note that this optimal money rule implies. i. Producing a deflation at the optimal rate (the rate of time preference) eliminates the distortion of the labor supply decision and brings about an optimal allocation of resources. interest rates. If alternative assets bear a higher real return than money. the optimum is characterized by the firstorder condition (8.27). we therefore require that " # γn∗ u0 (γn∗ ) βEt − n∗ v0 (n∗ ) = 0. the model has some problems in its ability to fit the facts. from (8.29). (8. the nominal interest rate is zero and the cash-in-advance constraint does not bind.e. a constant.40) this requires that ∗ θt+1 = β.8. we want t ∗ to determine the θt which will imply that nt = n∗ is a competitive equilibrium outcome for this economy. The optimal money growth rule in (8. The first problem is that the velocity of money is fixed in this model. From (8. There are at least two straightforward .e. ∗ θt+1 and from (8. then the representative consumer economizes too much on money balances relative to the optimum.. in the long run and over the business cycle. for all t. the money supply decreases at the discount rate.41) i. Letting n∗ denote t the optimal choice for nt . as does a positive nominal interest rate. PROBLEMS WITH THE CASH-IN-ADVANCE MODEL 115 but this breaks down into a series of static problems. and real activity. t t and this then implies that n∗ = n∗ . that St = 1 for all t. In this model. 8. a binding cash-in-advance constraint represents an inefficiency.4. but is highly variable in the data.41) is referred to as a “Friedman rule” (see Friedman 1969) or a “Chicago rule.40) γu0 (γn∗ ) − v 0 (n∗ ) = 0. Now.

counterfactual responses to surprise increases in money growth are predicted. Work by Lucas (1990) and Fuerst (1992) on a class of “liquidity effect” models. variability in inflation causes substitution between cash goods and credit goods. A second approach is to change some of the timing assumptions concerning transactions in the model. which in turn leads to variability in velocity. then there could not be an equilibrium with a positive nominal interest rate. and because it requires the modeler to define at the outset what money is. in versions of this type of model where money growth is serially correlated (as in practice). as a positive nominal interest rate represents a profit opportunity for private issuers of money substitutes. and velocity is variable. Implicit in the model is the assumption that private agents cannot produce whatever it is that satisfies cash-in-advance. surprise increases in money growth appear to generate short run increases in output and employment. The model is silent on what the objects are which enter the cash-in-advance constraint. money growth rates are positively serially correlated. which are versions of the cash-in-advance approach. A CASH-IN-ADVANCE MODEL means for curing this problem (at least in theory). Given this. Svennson (1985) assumes that the asset market opens before the current money shock is known. Hodrick. and. Here. and a short run decrease in the nominal interest rate. But this will imply (in this model) that labor supply falls. high money growth is expected tomorrow. Kocherlakota. Because the model is not explicit about the underlying restrictions which support cash-in-advance. given anticipated inflation. the cash-in-advance approach is virtually useless for studying sub- . For example. Empirically. the nominal interest rate rises. Empirically. A third problem has to do with the lack of explicitness in the basic approach to modeling monetary arrangements here. the cash-in-advance constraint binds in some states of the world but not in others. can obtain the correct qualitative responses of interest rates and output to money injections. If they could. In this context. Thus. Another problem is that. The first is to define preferences over “cash goods” and “credit goods” as in Lucas and Stokey (1987). and Lucas (1991) show that these models do not produce enough variability in velocity to match the data. output falls.116 CHAPTER 8. if there is high money growth today. cash goods are goods that are subject to the cash-in-advance constraint. neither of these approaches works empirically. However.

” Journal of Political Economy 110. 1991.” Journal of Monetary Economics 29. “The Inflation Tax in a Real Business Cycle Model. 1997. “Money and Exchange Rates in the Grossman-Weiss-Rotemberg Model.” Journal of Political Economy 97. 1969. There are approaches which model monetary arrangements at a deeper level. “Money. 1989. Cooley. and interest rates. Interest Rates. . P. 1992. F. “The Variability of Velocity in Cash-in-Advance Models. Loanable Funds. T. 619-40. N. Alvarez. T. and Hansen. and Kehoe. 927-954. Kiyotaki. 2002. 733-748. Atkeson. it is very difficult to argue that consumption expenditures during the current quarter (or month) are constrained by cash acquired in the previous quarter (or month). Aldine Publishing. and Real Activity. The Optimum Quantity of Money and Other Essays. “Liquidity.” Journal of Political Economy 99. D. prices. and Atkeson. REFERENCES 117 stitution among money substitutes and the operation of the banking system. Kocherlakota. Friedman. Fuerst.8. Hodrick. and Lucas. “On Money as a Medium of Exchange. Clearly. A last problem has to do with the appropriateness of using a cashin-advance model for studying quarterly (or even monthly) fluctuations in output. M.” Journal of Monetary Economics 40. Chicago.. A. such as in the overlapping generations model (Wallace 1980) or in search environments (Kiyotaki and Wright 1989). but these approaches are not easily amenable to empirical application. 1989. A. 8.5 References Alvarez. N. R. 358-384. 3-24. 73-112.” American Economic Review 79. given the low cost of visiting a cash machine or using a credit card.. and Exchange Rates with Endogenously Segmented Markets.5. and Wright. F. G. R.

” in Models of Monetary Economies. Federal Reserve Bank of Minneapolis. “Equilibrium in a Pure Currency Economy. “Interest rates and Currency Prices in a Two-Country World. N. “Money and Interest in a Cash-InAdvance Economy. Lucas. MN. 1990.118 CHAPTER 8. 1987.” in Models of Monetary Economies. “Liquidity and Interest Rates. Federal Reserve Bank of Minneapolis. A CASH-IN-ADVANCE MODEL Lucas.. 491-514. 237-264. Lucas.” Journal of Monetary Economics 10. N. Lucas. R.” Econometrica 55. Wallace. 1980. Kareken and Wallace eds. 1980. 1982. and Stokey. . Kareken and Wallace eds.. “The Overlapping Generations Model of Fiat Money. MN. 335-359. R. Minneapolis. R.” Journal of Economic Theory 50. Minneapolis. R.

suppose a world in which there is a finite number of different goods. but from its wide acceptability in transactions. and a unit of account. trades of goods for goods. i. In order to directly obtain the good she wishes. and agents will search. money is an asset. money is a unit of account in that virtually all contracts are denominated in terms of it.e. i. a long time for trading partners. its value is not derived solely from its intrinsic worth. and no trades of this type can take place. on average. Also suppose that all trade in this economy involves barter. trading will be a random and time-consuming process.e. money has been viewed as having three functions: it is a medium of exchange. it is necessary for a particular agent to find someone else who has what she wants. The key elements of Jevons’s story are that economic agents are specialized in terms of what they produce and what they consume. In the worst possible scenario. there is a double coincidence of wants. For example.e. A trade can only take place if that other person also wants what she has. Money is a medium of exchange in that it is an object which has a high velocity of circulation. Finally. At best. 119 .Chapter 9 Search and Money Traditionally. i. which is a single coincidence of wants. there is an absence of double coincidence of wants. Jevons (1875) provided an early account of a friction which gives rise to the medium-of-exchange role of money. and that it is costly to seek out wouldbe trading partners. a store of value. and each person produces only one good and wishes to consume some other good. It is hard to conceive of money serving its role as a medium of exchange without being a store of value.

and is useful for studying commodity monies. and a given agent can produce only one of the goods in the continuum. two single coincidences on average occur much sooner than one double coincidence. but is not a very tractable model of fiat money. Kiyotaki and Wright’s model involves three types of agents and three types of goods (the simplest possible kind of absence of double coincidence model). Rather than having to satisfy the double coincidence of wants. which is intrinsically useless but difficult or impossible for private agents to produce.1 The Model ∞ X t=0 There is a continuum of agents with unit mass. and . but the more recent monetary search literature begins with Kiyotaki and Wright (1989). agents meet pairwise and at random. the probability that her would-be trading partner produces a good that she likes to consume is x. An agent gets zero utility from consuming her production good. The model we study in this chapter is a simplification of Kiyotaki and Wright (1993). where 0 < β < 1. and then find an agent who has what she wants. and u(·) is an increasing function. or it could be fiat money. where symmetry is exploited to obtain a framework where it is convenient to study the welfare effects of introducing fiat money. an agent now only needs to find someone who wants what she has. One of the first models of money and search is that of Jones (1976). Each period. 9. ct denotes consumption. For a given agent. each having preferences given by E0 β t u(ct ). The above story has elements of search in it. If money is accepted by everyone. When there is a large number of goods in the economy. There is a continuum of goods. then trade can be speeded up considerably. purchasing their consumption good with money. selling her production for money. SEARCH AND MONEY Suppose now that we introduce money into this economy. so it is not surprising that the search structure used by labor economists and others would be applied in monetary economics. This money could be a commodity money. which is valued as a consumption good.120 CHAPTER 9.

where 0 ≤ π ≤ 1. and let π0 denote the probability with which the agent accepts money. From an individual agent’s point of view. and we let Vg denote the value of holding a commodity. Given symmetry. which occurs with probability x2 . but the agent with the commodity may or may not want to accept money.e. and all goods are stored at zero cost. and assume that the utility from consuming a good one does not like is zero. In a steady state.2 Analysis We confine the analysis here to stationary equilibria. Any intertemporal trades or gift-giving equilibria are ruled out by virtue of the fact that no two agents meet more than once. and Vm the value of holding money at the end of the period. but in either case they each end the period holding money.9. and a fraction M of the population is endowed with one unit each of this stuff in period 0. since both are indifferent. There is a good called money. If two agents meet and one has money while the other has a commodity. so it is possible to throw money (or anything else) away. all agents are holding one unit of some good (ignoring uninteresting cases where some agents hold nothing). If two agents with money meet. All goods are indivisible. equilibria where agents’ trading strategies and the distribution of goods across the population are constant for all t. For convenience. they will trade only if there is a double coincidence of wants. it is as if there are were only two goods. If two agents with commodities meet. ANALYSIS 121 the probability that she produces what her would-be trading partner wants is also x. we . being produced and stored in one-unit quantities. Then. The fraction of agents holding money is μ. they may trade or not. i. Free disposal is assumed. let u∗ = u(1) denote the utility from consuming a good that the agent likes.2. let π denote the probability that other agents accept money. the agent with money will want to trade if the other agent has a good she consumes. An agent can store at most one unit of any good (including money). 9. and because agents have no knowledge of each others’ trading histories. and the fraction holding commodities is 1−μ.

The first we .e. Now.1). and meets a commodity-holder with probability 1 − μ. Similarly. even though one agent is indifferent between trading and not trading. as we did in 1 the previous chapter. commodity equilibria. Trade with a commodity-holder occurs with probability xπ. and the commodity-holder accepts money with probability π. an agent holding money meets another agent holding money with probability μ. Now.3) (9.2). we will ignore equilibria where agents accept commodities in exchange that are not their consumption goods. and thus a commodity holder would strictly prefer not to trade for a commodity she does not consume. (9. If the agent trades.122 CHAPTER 9.4) rVm = (1 − μ)xπ(u∗ + Vg − Vm ). The money-holder will want to trade with probability x. With probability 1 − μ the agent meets another commodity-holder. there are potentially three types stationary equilibria.1] [π 0 Vm + (1 − π 0 )Vg ] + μ(1 − x)Vg +(1 − μ) [x2 (u∗ + Vg ) + (1 − x2 )Vg ] ) . It is convenient to simplify the Bellman equations. One type has π = 0.2) to get rVg = μx max π 0 (Vm − Vg ) + (1 − μ)x2 u∗ . 0 π ∈[0.2) Vm = β {μVm + (1 − μ) [xπ(u∗ + Vg ) + (1 − xπ)Vm ]} .1] (9. as there is a single coincidence with probability x. SEARCH AND MONEY can write the Bellman equations as Vg = β ( μx maxπ0 ∈[0. and trade takes with probability x2 . Provided ε is very small. the agent chooses the trading probability π 0 to maximize end-of-period value. In (9. they trade. and one has π = 1. and manipulating (9. i. defining the discount rate r by β = 1+r .1) (9. the analysis does not change. one has 0 < π < 1. as in Kiyotaki and Wright (1993) by assuming that there is a small trading cost.1) and (9. if two commodity-holders meet and there is a single coincidence. ε > 0. and if the moneyholder wishes to trade. she consumes and then immediately produces again. In these equilibria. in (9. It is easy to rule these equilibria out. an agent with a commodity at the end of the current period meets an agent with money next period with probability μ.

Further. indexed by μ ∈ (0. in conjunction with the assumptions about the inventory technology. Next. Thus. implies that less consumption takes place in the aggregate when money is introduced. Suppose first that π = 0. we solve (9. note that all money-holders are indifferent between throwing money away and producing. the expected utility of each agent in equilibrium is Vg = x2 u∗ .2. From (9. Then. so introducing money in this case does nothing to improve trade.3) we must have Vm = Vg .5) Next. Then. Conjecturing that this is so. This then gives expected utility for all agents in the stationary equilibrium Vm = Vg = (1 − μ)x2 u∗ . so we have μ = 0.7) . so we must have Vm ≥ Vg . there is a continuum of equilibria of this type. so for the mixed strategy to be optimal. the fact that some agents are holding money.9. and that expected utility is decreasing in μ.3) and (9. from (9. r(r + x) (9.4) for Vm and Vg to get Vg = (1 − μ)x2 u∗ [μ(1 − x) + r + x] . in the mixed strategy monetary equilibrium. Here. note. it must be optimal for the commodity-holder to choose π 0 = π = 1. r (9.6) Now. and the latter two are monetary equilibria. In equilibrium we must have π 0 = π.4). This is due to the fact that. an agent holding money would never get to consume. from (9.3). from (9.3) and (9. that all agents are worse off in the mixed strategy monetary equilibrium than in the non-monetary equilibrium.5) and (9. ANALYSIS 123 can think of as a non-monetary equilibrium (money is not accepted by anyone). consider the mixed strategy equilibrium where 0 < π < 1.6). money is no more acceptable in exchange than are commodities (π = x). and anyone holding money at the first date would throw it away and produce a commodity. In addition. r (9. we then must have π = x. and holding their money endowment. M]. consider the equilibrium where π = 1.

Thus our conjecture that π 0 = 1 is a best response to π = 1 is correct. If money is allocated to agents at random at t = 0. and we will have μ = M. it is useful to consider what welfare is in the monetary equilibrium with π = 1 relative to the other equilibria. then introducing 2 ∗ . if x ≥ 1 . 2 i.7) and (9. Now. which is identical to welfare in the non-monetary equilibrium. Suppose that we imagine a policy experiment where the monetary authority can consider setting M at t = 0. this is the expected utility of each agent before the money allocations occur. Setting μ = M in (9. the optimal quantity of money is M ∗ = 0.124 Vm = and we have CHAPTER 9. That is. SEARCH AND MONEY (1 − μ)xu∗ [−(1 − μ)x(1 − x) + r + x] .8). for M = 0. and calculating W. we obtain dW xu∗ = [1 − 2x + 2M(−1 + x)]. W = x ru . then introducing any quantity of money reduces welfare. r(r + x) Vm − Vg = (9. dM r d2 W = 2(−1 + x) < 0. weighted by the population fractions.e. dM 2 Thus. i. we will use as a welfare measure W = (1 − M)Vg + MVm . Here.e. This does not correspond to any real-world policy experiment (as money is not indivisible in any essential way in practice). as all agents with a money endowment will strictly prefer holding money to throwing it away and producing. as should be the case. (9. Differentiating W with respect to M. the expected utilities of the agents at the first date.8) (1 − μ)x(1 − x)u∗ >0 r+x for μ < 1. but is useful for purposes of examining the welfare effects of money in the model. if the absence of double coincidence of wants problem is not too severe. we get (1 − M)xu∗ W = [x + M(1 − x)] .9) r Note that.

Note that from (9. However. a change in M is essentially a meaningless experiment. 9. . The model has been extended to allow for divisible commodities (Trejos and Wright 1995. if there is some knowledge of a would-be trading partner’s history. we need to track the whole distribution of money balances across the population. In indivisible-money search models. for a given μ. such as changes in the money growth rate which would affect inflation. though this has been done in computational work (Molico 1997). Weber and Wright 1998). we need a sufficiently severe absence of double coincidence problem before welfare improves due to the introduction of money.” Shi (1996) also studies a monetary search model with credit arrangements of a different sort. and a role for money arising from informational frictions (Williamson and Wright 1994).3.9.3 Discussion This basic search model of money provides a nice formalization of the absence-of-double-coincidence friction discussed by Jevons.9). it is impossible to consider standard monetary experiments. Shi 1995). If x < 1 . it is possible to have credit-like arrangements. then welfare is maximized 2 1−2x for M ∗ = 2(1−x) . A remaining problem is that it is difficult to allow for divisible money. While credit is ruled out in the above model. DISCUSSION 125 money reduces welfare more by crowding out consumption than it increases welfare by improving trade. Thus. If money is divisible. Further. if money is not divisible. Kocherlakota and Wallace (1998) and Aiyagari and Williamson (1998) are two examples of search models with credit arrangements and “memory. it has been used to address historical questions (Wallace and Zhou 1997.6) and (9. Velde. which is analytically messy. even if no two agents meet more than once. welfare is higher in the pure strategy monetary equilibrium than in the mixed strategy monetary equilibrium.

and Williamson. forthcoming. 555-572. Kocherlakota. R.” working paper. “Optimal Allocations with Incomplete Record-Keeping and No Commitment. “Credit and Money in a Search Model with Divisible Commodities. Molico. N. 1875.” Journal of Political Economy 97. N. Shi. 927-954.” forthcoming. “A Search-Theoretic Approach to Monetary Economics 83. and Wright. and Wright. 1997.” Review of Economic Studies 63.” forthcoming. “The Distributions of Money and Prices in Search Equilibrium. 757-775. “Search. Jones. and Wright. 63-77.4 References Aiyagari. 1976. Appleton. and Prices. 1998. R.126 CHAPTER 9. Shi. and Zhou. 1997. R. “Credit in a Random Matching Model with Private Information. F.” Journal of Monetary Economics 40. Kiyotaki. and Wright.” Journal of Political Economy 103. Money and the Mechanism of Exchange. 1993. “A Model of Commodity Money. A. R. Kiyotaki. . 1995. Weber. Bargaining. Money. “ A Model of a Currency Shortage.” Review of Economic Dynamics. London. N. N. Trejos. S. W. 1989.. 627-652. 1998. and Wallace. R. Velde. W. Journal of Economic Theory. 467-496. University of Pennsylvania. 1996. N. S. 1995. “The Origin and Development of Media of Exchange.” Journal of Economic Theory 67. Wallace. 1998. Review of Economic Dynamics. S. Jevons. S. SEARCH AND MONEY 9. “On Money as a Medium of Exchange. with Applications to Gresham’s Law and the Debasement Puzzle. M. R. 118-141. “Money and Prices: A Model of Search and Bargaining.” Journal of Political Economy 84.

4.” American Economic Review 84. and Wright. REFERENCES 127 Williamson. 1994. .9. S. “Barter and Monetary Exchange Under Private Information. R. 104123.

128 CHAPTER 9. SEARCH AND MONEY .

as we will see. this may seem silly if taken literally. for example). we studied Peter Diamond’s overlapping generations model of growth. in Chapter 2. the overlapping generations model of money includes an explicit representation of the absence of double coincidence 129 . who did not take it very seriously. In fact. Earlier.Chapter 10 Overlapping Generations Models of Money The overlapping generations model of money was first introduced by Samuelson (1956). money is used in the overlapping generations environment because it overcomes a particular friction that is described explicitly in the model. In the simplest overlapping generations models. In terms of how money works in real economies. in the late 1970s and early 1980s (see Kareken and Wallace 1980. the friction is that agents are finite-lived and a particular agent can not trade with agents who are unborn or dead. As in the search model of money. his coauthors and students. agents hold money in order to consume in their old age. which was an adaptation of Samuelson’s model used to examine issues in capital accumulation. and it was then used extensively by Neil Wallace. Samuelson’s monetary model was not rehabilitated until Lucas (1972) used it in a business cycle context. However. the overlapping generations friction should be interpreted as a convenient parable which stands in for the spatial and informational frictions which actually make money useful in practice. In this case. since the holding period of money is typically much shorter than thirty years.

the government budget constraint is pt (Mt − Mt−1 ) = Nt−1 τt . t t+1 where cs denotes consumption in period t by a member of generation t s. in terms of the period t consumption good (which will be the numeraire throughout). 3..c2 ) ∂c2 = 0. Assume that u(·. In period 1.1 The Model In each period t = 1. Money can be injected or withdrawn through lump-sum transfers to old agents in each period. 2. 10.c2 ) lim ∂c1 c2 →0 ∂u(c1 . OVERLAPPING GENERATIONS MODELS OF MONEY problem.. for c1 > 0. Assume that the population evolves according to Nt = nNt−1 . ct ). . These agents are collectively endowed with M0 units of fiat money. We will call agents young when they are in the first period of life. and that ∂u(c1 . who live for only one period. Each young agent receives y units of the perishable consumption good when young. . (10.. Letting τt denote the lump sum transfer that each old agent receives in period t. which is perfectly divisible. there are N0 initial old agents.c2 ) lim ∂c1 c1 →0 ∂u(c1 . and can not be privately produced. who are endowed with money). ·) is strictly increasing in both arguments. and letting Mt denote the money supply in period t. 3. and each old agent receives nothing (except for the initial old. Nt agents are born who are each twoperiod-lived. strictly concave and twice continuously differentiable. and old when they are in the second period. 2. for c2 > 0.130CHAPTER 10.1) for t = 1. and whose utility is increasing in period 1 consumption.. (10. which will guarantee that agents want to consume positive amounts in both periods of life.2) .. and ∂u(c1 .. where n > 0. intrinsically useless.c2 ) ∂c2 = ∞. An agent born in period t has preferences given by u(ct .

allocations that have the property that each generation born in periods t = 1..e. receives the same allocation. ct ) = (c1 . pt = 0.. i. Mt = zMt−1 . . . i.3) 10. with z > 0. PARETO OPTIMAL ALLOCATIONS 131 for t = 1. This planner faces the resource constraint Nt ct + Nt−1 ct−1 ≤ Nt y.5) using (10. t t (10..1) to get c2 (10. It is convenient to use the consumption good as the numeraire.2) and (10..e. 2...6) are feasible. We can then rewrite equation (10.. for t = 1. Note that the initial old alive in period 1 will then each consume c2 . further. where pt denotes the price of money in terms of the period t consumption good. c2 ) satisfying (10. where t t+1 c1 and c2 are nonnegative constants..2. Now. as we will want to consider equilibria where money is not valued. for all t. c2 ). suppose that there is a social planner that can confiscate agents’ endowments and then distribute them as she chooses across the population.2 Pareto Optimal Allocations Before studying competitive equilibrium allocations in this model. To that end.5) for t = 1. i. suppose that the planner is restricted to choosing among stationary allocations. 2. 3. 3. 3. so (10. 1 .3) imply that 1 pt Mt (1 − ) = Nt−1 τt .5) states that total consumption of the young plus total consumption of the old can not exceed the total endowment in each period. we wish to determine what allocations are optimal. Equation (10. . . n We will say that stationary allocations (c1 .e. 3.4) (10.1 Further. 2.6) c1 + ≤ y. we will assume that the money supply grows at a constant rate. 2... z (10.10.e. i. . or (ct . the inverse of the price level.

c2 ) and c2 ≥ c2 . i. there is some alternative allocation which satisfies (10. Thus.e.. Then. (c∗ . c2 ) such that c ˆ u(ˆ1 . OVERLAPPING GENERATIONS MODELS OF MONEY Definition 3 A Pareto optimal allocation. .c 1 2 To see this.. c2 ). 2. note that we take account of the welfare of the initial old agents. c2 ) c .. c∗ ) is the 1 2 solution to max u(c1 .6) where (10.7) is not satisfied. 3.10) . t ct t+1 = pt+1 mt + τt+1 (10. ct ≥ 0.6) with equality. the definition states that an allocation is Pareto optimal (within the class of stationary allocations) if it is feasible and there is no other feasible stationary allocation for which all agents are at least as well off and some agent is better off. ct ) (10. and selling them when old.6). Thus.7) 1 10. letting mt denote the nominal quantity of money acquired by an agent born in period t.132CHAPTER 10. note that. the Pareto optimal allocations satisfy (10. note first that any Pareto optimal allocation must satisfy (10. To determine what allocations are Pareto optimal. (c1 .3 Competitive Equilibrium A young agent will wish to smooth consumption over her lifetime by acquiring money balances when young. chosen from the class of stationary allocations. Here.6) and makes all agents better off.8) t t+1 subject to ct + pt mt = y. for any allocation satisfying (10.9) (10. c2 ) ≥ u(c1 . c∗ ) denote the stationary allocation that maximizes the 1 2 welfare of agents born in generations t = 1. and ct ≥ 0 to t t+1 solve max u(ct . Further. is feasible and satisfies the property that there exists no other feasible stationary allocation (ˆ1 . with at least one of the previous two c ˆ ˆ inequalities a strong inequality. subject to (10.6) with equality and c1 ≤ c∗ . (10. the agent chooses mt ≥ 0. let (c∗ .

0) and τt = 0 for all t.. 2. and (iii) is the market-clearing condition.. . COMPETITIVE EQUILIBRIUM 133 Definition 4 A competitive equilibrium is a sequence of prices {pt }∞ . and wishes to trade some of these for period t + 1 consumption goods. 1 2 Thus.. Note that there are two markets. a sequence of money t t+1 t=1 supplies {Mt }∞ .8) subject to (10.. . condition (ii) states that the sequence of money supplies and lump sum taxes satisfies the constant money growth rule and the government budget constraint. there always exists a non-monetary equilibrium. 3. for all t = 1.9) and (10. t=1 t=1 and a sequence of taxes {τt }∞ . t t+1 It is straightforward to verify that conditions (i)-(iii) in the definition of a competitive equilibrium are satisfied. However. i. condition (i) says that all agents optimize treating prices and lump-sum taxes as given (all agents are price-takers). . a competitive equilibrium where money is not valued and pt = 0 for all t. (ii) (10. a sequence of individual money demands {mt }∞ . 2. . 2. pt+1 . ct ) = t t+1 (c∗ . in the absence of money. and τt+1 . .. In the nonmonetary equilibrium. ct ) t=1 t t+1 and mt are chosen to solve (10. since it is Pareto dominated by the feasible stationary allocation (ct . t=1 a sequence of consumption allocations {(ct .10. That is.4). the market for consumption goods and the market for money. 10. given M0 . but Walras’ Law permits us to drop the market-clearing condition for consumption goods. an agent born in period t has period t consumption goods.3.10) given pt . It is also straightforward to show that the nonmonetary equilibrium is not Pareto optimal. ct ) = (y. 3. .. due to a type of absence-of-double-coincidence friction.3) and (10. c∗ ). . 3. for t = 1. ct )}∞ . there is no other agent who wishes to trade period t + 1 consumption goods for period t consumption goods.3.. (ct .1 Nonmonetary Equilibrium In this model.e. (iii) pt Mt = Nt pt mt for all t = 1. In the definition. which satisfies: (i) (ct . no trade can take place in this model.

Once we solve for {qt }∞ . pt+1 mt + τt+1 ). qt+1 n). at n −qt u1 (y − qt . pt+1 mt + τt+1 ) + pt+1 u2 (y − pt mt . Now. and the z sequence of consumptions is given by (ct . To solve for q.11) where ui (c1 . (10. pt+1 mt + τt+1 ) = 0. t=1 t=1 tN we can then work backward to solve for {pt }∞ . We can then Nt use this definition of qt .9) and (10. c2 ) denotes the first partial derivative of the utility function with respect to the ith argument. We will also suppose (as has to be the case in equilibrium) that taxes and prices are such that an agent born in period t chooses mt > 0. ct ) = (y − qt . t=1 where qt ≡ pt Mt is the real per capita quantity of money. qt+1 n) = 0 z (10. substitute for the constraints in the objective function to obtain max u(y − pt mt .11) to arrive.8) subject to (10. qt+1 n) + qt+1 u2 (y − qt .3. for all t. t=1 The sequence of taxes can be determined from τt = nqt (1 − 1 ). This competitive equilibrium has the property that qt = q. and conditions (ii) and (iii) in the definition of competitive equilibrium to substitute in the first-order condition (10. t t+1 One monetary equilibrium of particular interest (and in general this will be the one we will study most closely) is the stationary monetary equilibrium. assuming an interior solution. the first-order condition for an optimum is −pt u1 (y − pt mt . To simplify the problem (10.2 Monetary Equilibria We will now study equilibria where pt > 0 for all t. a constant. it proves to be convenient in this version of the model (though not always) to look for an equilibrium in terms of the sequence {qt }∞ . simply set qt+1 = qt = q in equation . Here. after some manipulation. m t and then.134CHAPTER 10. given that pt = qMtt .12) is a first-order difference equation in qt which can in principle be solved for the sequence {qt }∞ . OVERLAPPING GENERATIONS MODELS OF MONEY 10. given our assumptions on preferences. agents will choose an interior solution with strictly positive consumption in each period of life.12) Equation (10.10).

10. (y − qt )n (10. . 2. The consumption allocations are zy ny (ct . a fixed money supply is Pareto optimal.6). i. equation (10. qn) + u2 (y − q.13) z Now. Here.12) gives (after rearranging) qt+1 = 2 qt z 2 . so optimality here has nothing to do with what the inflation rate is.. i. and z ≤ 1 implies that (10.e. . 1+z ). so that consumption of the young increases with t t+1 the money growth rate (and the inflation rate). so the stationary monetary equilibrium is the unique monetary equilibrium in this case (though note that qt = 0 is still an equilibrium). note that. c2 ) = c1 + c2 . 3. then y−q = c∗ and qn = c∗ . Further. c2 ) = ln c1 + ln c2 . y − qt z Suppose first that u(c1 . since the stationary monetary equilibrium must satisfy (10.12) to obtain 135 n −u1 (y − q. EXAMPLES (10. independent of the population growth rate. 1 1 2 2 Alternatively. and consumption of the old decreases..12) gives y and solving for qt we get qt = 1+z . the agent faces intertemporal prices which are different from the terms on which the social planner can exchange period t consumption for period t + 1 consumption. If z > 1. qn) = 0 (10. Then.10. note that any z ≤ 1 implies that the stationary monetary equilibrium is Pareto optimal. due to market clearing. by virtue of the 1 2 fact that each agent is essentially solving the same problem as a social planner would solve in maximizing the utility of agents born in periods t = 1. equation (10. suppose that u(c1 .4 Examples qt 1 = .4.e. ct ) = ( 1+z . this implies that intertemporal prices are distorted. Thus. z = 1 implies that the stationary monetary equilibrium is Pareto optimal.14) .7) holds in the stationary monetary equilibrium. if z = 1. Note that the rate z of inflation in the stationary monetary equilibrium is n −1.

(10. Note that the stationary monetary equilibrium satisfies z the quantity theory of money. There are many other types of multiple equilibria that the overlapping generations model can exhibit. The model has been criticized for being too stylized.6 References Azariadis. . i. some see the existence of multiple equilibria as being undesirable. the nonstationary monetary equilibria do not have this property. it is easy to integrate other features into the model. n+z2 ). i. However. for doing empirical work the interpretation of period length is problematic. 10.5 Discussion The overlapping generations model’s virtues are that it captures a role for money without resorting to ad-hoc devices. indexed by q1 ∈ (0. though some in the profession appear to think that the more equilibria a model possesses. OVERLAPPING GENERATIONS MODELS OF MONEY Now. the better. in some versions of the model). and it is very tractable. There also exists a continuum ny of equilibria.e. 1981. one of which is the stationary ny monetary equilibrium where qt = n+z2 . and Sargent 1987). since the agents in the model need only solve two-period optimization problems (or three-period problems.1 10. these are nonstationary monetary equilibria where there is convergence to the nonmonetary equilibrium in the limit. including “sunspot” equilibria (Azariadis 1981) and chaotic equilibria (Boldrin and Woodford 1990). Each of these equilibria has the property that limt→∞ qt = 0. Also. “Self-Fulfilling Prophecies.136CHAPTER 10. such as credit and alternative assets (government bonds for example) by allowing for sufficient within-generation heterogeneity (see Sargent and Wallace 1982. C. Bryant and Wallace 1984.e. so that increases in the money growth rate are essentially reflected onefor-one in increases in the inflation rate (the velocity of money is fixed at one). 380-396.14) has multiple solutions.” Journal of Economic Theory 25. Further. in that the inflation rate is n − 1.

1972. 467-482. Lucas. “Equilibrium Models Displaying Endogenous Fluctuations and Chaos: A Survey. T.” Journal of Political Economy 66. “Expectations and the Neutrality of Money. and Wallace. N. 1212-1236. Federal Reserve Bank of Minneapolis. and Woodford. R. Samuelson. P. 279-288. REFERENCES 137 Boldrin.10. Harvard University Press. 1990. M. “An Exact Consumption Loan Model of Interest With or Without the Social Contrivance of Money. 1956. and Wallace. MA. Sargent. Minneapolis.” Journal of Political Economy 90. N. “The Real Bills Doctrine vs.6. . the Quantity Theory: A Reconsideration. “A Price Discrimination Analysis of Monetary Policy. and Wallace. 1984. M.” Journal of Monetary Economics 25. Models of Monetary Economies. 1987. 103-124. 1982.” Review of Economic Studies 51. J. Dynamic Macroeconomic Theory. N.” Journal of Economic Theory 4. Cambridge. 189-222. Bryant. Kareken. T. Sargent. MN. J. 1980.

Sign up to vote on this title
UsefulNot useful