This action might not be possible to undo. Are you sure you want to continue?

Uploaded by pilotofsecret

0.0 (0)

, Volume 57, Issue 3 (Jul., 1990),403-414.

STOR

Your use of the JSTOR database indicates your acceptance of JSTOR' s Terms and Conditions of Use. A copy of JSTOR's Terms and Conditions of Use is available at http://www.jstor.org/aboutiterms.html, by contacting JSTOR atjstor-info@umich.edu, or by calling JSTOR at (888)388-3574, (734)998-9101 or (FAX) (734)998-9113. No part of a JSTOR transmission may be copied, downloaded, stored, further transmitted, transferred, distributed, altered, or otherwise used, in any form or by any means, except: (1) one stored electronic and one paper copy of any article solely for your personal, non-commercial use, or (2) with prior written permission of JSTOR and the publisher of the article or other text. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. The Review of Economic Studies is published by The Review of Economic Studies Ltd.. Please contact the publisher for further permissions regarding the use of this work. Publisher contact information may be obtained at http://www.jstor.org/joumals/resl.html.

The Review of Economic Studies ©1990 The Review of Economic Studies Ltd.

JSTOR and the JSTOR logo are trademarks of JSTOR, and are Registered in the U.S. Patent and Trademark Office. For more information on JSTOR contactjstor-info@umich.edu. ©2000JSTOR

http://www .jstor.org/ Fri Oct 612:13:192000

Barro (1985.' In 1. Los Angeles. Government debt is used strategically by each policymaker to influence the choices of his successors. NBER and CEPR First version received December 1988. The paper abandons the assumption that fiscal policy is set by a benevolent social planner who maximizes the welfare of a representative consumer.) This paper considers an economy in which policymakers with different preferences alternate in office as a result of elections. final version accepted September 1989 (Eds. this theory may not provide a complete explanation of two recent facts: (a) the recent rapid accumulation of government debt in several industrialized countries. 1. for instance. If different policymakers disagree about the desired composition of government spending between two public goods. (b) the large variation in the debt policies pursued by different countries with similar economic conditions. This paper suggests a positive theory of government debt policy which should provide useful insights in explaining these facts. during peacetime.00 © 1990 The Review of Economic A Positive Theory of Fiscal Deficits and Government Debt ALBERTO ALESINA Harvard University. much of it by James Buchanan and his associates. Instead. In their model. NBER and CEPR and GUIDO TABELLINI University of California. However. The equilibrium level of debt is larger the larger is the degree of polarization between alternating governments and the less likely it is that the current government will be re-elected.Wagner (1977) and Brennan-Buchanan (1980). unlike in ours. 403-414 Studies Limited 0034-6527/90/00250403$02. including the United States. as in Barro (1979) and Lucas-Stokey (1983). 403 . public debt is modelled as a means of distributing tax distortions over time. 1987) has shown that this normative theory of fiscal policy can explain quite well the behaviour of public debt in the United States (at least up to the late seventies) and in the United Kingdom. Most of this literature. Variants of this explanation of why governments run fiscal deficits can be found. in Buchanan. however. debt accumulation is higher than it would be with a social planner. INTRODUCTION Budget deficits and debt accumulation can serve two purposes: they provide a means of redistributing income over time and across generations. Cukierman-Meltzer (1989) provide a rational politico-economic model of public debt. public debt is used for intergenerational transfers and taxes are non-distortionary. and they serve as a means of minimizing the deadweight losses of taxation associated with the provision of public goods and services. This paper focuses on the latter issue. the economy exhibits a deficits bias. is based upon the somewhat questionable notions of "fiscal illusion" and voters' irrationality. A large literature on the political economy of budget deficits already exists. we consider an economy with two policymakers with different objectives alternating in office as a result of elections. 1986. Thus. that is.Review of Economic Studies (1990) 57.

the equilibrium level of public debt tends to be larger: (i) the larger is the degree of polarization between alternating governments. and individuals act rationally and with full information as economic and political agents. in the presence of disagreement between current and future policymakers. In other words. the politicians' time horizon and discount factor coincide with those of the economy. According to the results of our model. Section 3 analyzes its static properties. They are all born at the beginning of period zero and have the same time horizon. (ii) the more likely it is that the curent governmentwill not be re-appointed. the equilibrium of our model generally exhibits a bias towards budget deficits. by contrast. at least three times continuously differentiable. Individual i has the following objective function (intratemporal separability serves only to simplify the algebra): 1> 8 >0. acting as consumers. that may generate a deficit bias in democracies. whereas we concentrate on disagreement about its composition. thus they assume an exogenously given world interest rate. v(·) and h(·) are continuous. Individuals differ only in their preferences for the two public goods. (iii) they do not explicitly consider voting behaviour. . while we have a probabilistic change of government and we study both a two-period and an infinite-horizon model. their disagreement is parameterized by the coefficient £:ri. workers and voters. thus the interest rate is endogenously determined. Nonetheless. Persson-Svensson (1989) have independently developed a related model. (1) where c is private consumption. 2. The functions u(·). rather than their myopia. Section 7 discusses some extensions and Section 8 concludes. More generally. Section 6 studies voting. Section 5 characterizes the dynamic economic equilibrium. As a result. THE MODEL We consider a closed economy composed of a large number of atomistic individuals. The outline of the paper is as follows: the model is presented in Section 2. and 8 is the discount factor. The optimal and time-consistent fiscal policies of an hypothetical social planner are described in Section 4.404 REVIEW OF ECONOMIC STUDIES this paper. our paper suggests that differences in political institutions can contribute to explain the variance in the debt policies pursued by different countries. (ii) they consider a two-period model in which the current government is sure that it will not be reappointed. or by the same country at different points in time. 2. g and f are two different public goods in per capita terms. The basic insight of the paper is that. strictly increasing and strictly concave. while we develop a voting equilibrium compatible with the economic equilibrium: and (iv) they consider a small open economy. the equilibrium stock of public debt tends to be larger than it is socially optimal. Both the finite-horizon case (two periods) and the infinite-horizon case are studied. Eo is the expectation operator conditional on the information available at time o. Their paper differs from ours in the following respects: (i) they focus on the disagreement about the level of public expenditures. public debt is used strategically by each government to influence the choices of its successors? Disagreement amongst alternating policymakers and uncertainty about who will be appointed in the future prevent the current government from fully internalizing the cost of leaving debt to its successors. it is the citizens' disagreement. x is leisure. (iii) the more rigid downward is public consumption. we instead consider a closed economy.

of different groups of consumers. i. (2) where q.' With identical tax rates for all consumers. We view these parties as the representatives of different "pressure groups". but cannot be addressed in a "real" model such as the present one. T. (See also footnote 11).)]}. Electoral results are uncertain: party D is elected with probability P and 3. q. The parties' preferences are as follows (the superscripts identify the party): WD WR = = Eo{I:=o 8'[u(c. to be shared between labour and leisure. The government can impose a proportional tax on labour income.)]}. a period is defined as a term of office. The general results are available from the authors. the tax rate and the amount of borrowing (lending) from the consumers. + v(x.Tabellini (1989) present a two. as in this model. is the present value at time zero of one unit of output at time t. in each period the government chooses the level and the composition of public consumption. Cases of default "de jure" are relatively uncommon. Our results generalize to arbitrary values of initial government debt. this role of the maturity structure concerns the strategic interaction between the government and the rest of the economy...) Eo{I:=o 8'[u(c. and qo = 1.). party D is identified with the consumer (or "constituency") with = 1. denoted D and R.party open economy model in which default is explicitly considered. can hold office. Cases of "de facto" default by means of inflation are much more common.)(I-x. and to simplify the analysis. As discussed in Chapter VIII of Persson and Tabellini (1990). which is the focus of this paper. Two "political parties". Elections are held at the beginning of each period. and not the interaction between alternating governments with different preferences. .) + h(f. Moreover. For these reasons. the superscript i on c and x can be dropped since all the consumers choose the same level of private consumption and leisure. Tabellini (1989) studies a model in which redistributional concerns induce the voters to repay the government debt even if default is explicitly allowed. i.) + v(x. Labour can be transformed one-for-one into non-storable output. The government can issue only fixed interest debt with one-period maturity." There is no default risk: each government is committed to honouring the debt obligations of its predecessors. (Xi (Xi (4) (5) Thus.. = «: qi being the inverse of the gross real interest rate in period i.) + h(g. With no loss of generality we assume throughout the paper that in period zero consumers do not hold any government debt. this result does not survive if government spending is determined endogenously. The preferences of the parties do not change over time and a prohibitive barrier prevents the entry of a third party. the parties' difference in preferences is only about the composition of public consumption. g and f Thus. 7" identical across consumers. Thus. 1. . . the intertemporal budget constraint of the consumer is: I:=oq. Lucas-Stokey (1983) show that the maturity structure of government debt can be used to ensure time-consistency of the socially optimal policy in a model with exogenous government spending.e. and party R is identified with the consumer (or "constituency") with = O.ALESINA & TABELLINI DEFICITS AND DEBTS 405 Each consumer is endowed with one unit of time. This assumption simplifies the algebra: Section 7 shows that our results generalize to the case in which both parties care about both public goods. Thus.period debt.e. 4. (3) ru. we only consider one.c.~I:=oq(1-7. Alesina. At no cost the government can transform the output produced by the private sector into the two non-storable public goods.' The resource constraint (in per-capita terms) is given by: t: 0. but with different weights. Since all consumers make identical choices regarding consumption and leisure. 5.

1988b). Is.) = x.406 REVIEW OF ECONOMIC STUDIES party R with probability 1 . These two results.). for a given deficit.} as given. Inspection of these maximization problems immediately yields the following results: (i) party D supplies only good g and party R only good 1. under the constraints given by (3). The "representative consumer" (for the purpose of private economic decisions) solves (2) and (3) taking the stochastic processes that determine {T. Since income taxes are the same for every consumer.' x. )]. In (7).) + h(l- c. TAXES AND PUBLIC SPENDING FOR A GIVEN DEFICIT No government can bind the taxation and expenditure policies of its successors. C. [u(c. (7) where u. Each government maximizes its own objective function. and Vx denote the derivative of u( . . hence proper sub-games do not exist. private consumption and leisure are identical under either party. where b. Both parties could benefit by agreeing to compromise to a certain constant composition of public spending. (6) and (7). The model is solved by means of dynamic programming. ) and v( . For expositional purposes P is temporarily assumed to be an exogenous constant: in Section 6 we complete the model with the endogenous determination of the political equilibrium. to insure that the solution is a sequentially rational equilibrium of the game in which each government plays against its successors and "against" the private sector.x.x. This point is further discussed in Persson.7 (ii) for a given fiscal deficit.)(1. .b. b. enable us to rewrite the static optimization problem faced by the party in office as follows. Max . the expectation reflects the uncertainty of consumers about the future tax policy. thus.T.)ue( c. although on different goods. and the intertemporal problem of choosing the size of the deficit. (8) (c.Tabellini (1990). ) ~ O. it is convenient to separate the government's optimization problem in two stages: the intra-period problem of choosing taxes and public consumption for a given fiscal deficit. (9) 6. 7. Throughout the paper we disregard the possibility of "cooperation" between the two parties. The game is partially anonymous because only aggregate variables and policy variables can be observed.(1. is government debt at the beginning of period t. 3. This agreement could be sustained as a sequentially rational equilibrium by means of trigger strategies as described in a different context by Alesina (1987." Since the government objective function is time-separable. ) vAx. regardless of whether the successor belongs to the same party or not. Sequential rationality is the analogue of sub-game perfection in partially anonymous policy games of this kind. both parties choose the same tax rate and the same level of public consumption.P.l and {j.) = vAx. ) with respect to their arguments (the arguments of the functions u(·) and v(·) will be omitted when there is no possibility of confusion). The only way in which the policy of the current administration can influence the actions of its successors is through the law of motion of public debt. the solution of this problem is characterized by: (6) uAc. .) + 8uA c'+I)bt+l . either (4) or (5).) + v(x. subject to: H( b'+I.}. together with the resource constraint (3).

has been obtained by substituting (6) and (7) into a dynamic form of (2). The second-order conditions are assumed to be satisfied.+l' RN(b" b. and all values of c and x dated t + 1 or later. For example it is shown that c* and x* are increasing.+l)) (11) Re(b" b. to eliminate T.. ).h(g*(b" b'+l))' Section 2 of the appendix proves the following: Lemma 1.+) == f*(b" b.. and b'+l.-b. affecting x and c indirectly. The second-order sufficient conditions are. 10. x*. All our results generalize to the case of a downward-sloping labour supply at the optimum if a very weak sufficient condition is satisfied (details are available). FISCAL POLICY WITH A SOCIAL PLANNER debt. starting with (a). and the constraints (3) and (9). the government does not choose x and c directly: it chooses taxes and public spending.+l) denote the utility function of either party if not in office in period t: r RN (b./H/ax2:. The expectations operator E. as noted in the text. 8.ALESINA & TABELLINI DEFICITS AND DEBTS 407 This problem is solved for given b. (10).. This assumption is sufficient (but not necessary) to prove some of the qualitative properties of c*. 9.0. v( ." Needless to say." The first-order condition.+l) is strictly increasing in b. We start A social planner has two characteristics: (a) he is "reappointed" with probability 1 in each period. With a given end of period debt (b'+l). . which implies assumptions about the third derivatives of u(·) and v(·). appearing in (2) has been omitted since. Let us indicate this function for the party in office as Re(b" b. we consider the effects of these two characteristics separately. say.+l) . in b. b. a2H/ac2:.0.. and g* and f* decreasing. and strictly concave in b. Throughout the paper we assume that the optimum is an interior point of the feasible set. Let RN (b. The condition for an upward-sloping labour supply function is: uc+ucc(c. (9). The intuition is that if b. 4. b.. the government is forced to reduce public consumption and to raise taxes.. The first-order conditions (if. (b) he adopts as his preferences a weighted average of the preferences of the citizens.+l) is strictly decreasing in b. We now turn to the choice of the optimal time path for government with the case in which policy is chosen by a social planner. f* and g*. implicitly define the optimal private and public consumption and leisure choices in period t as a function of b.+l)' With virtually no loss of generality we shall assume that at the optimum the labour supply function is upward sloping. party D is in office) imply: (10) where Hi denotes the derivative of H with respect to the variable i. since g* = and both parties choose the same tax rate.. and b. which are indicated as c*(b" bt+l). and b. The private sector responds with an increase of its consumption of both leisure and output. g*(b" b.)~O.+l' The government budget constraint. in addition to the strict concavity of u( . Re(b" b. increases more debt has to be serviced in period t.+l) = u(c*(b" = bt+l)) + v(x*(b" b. ): . Both R" and RN are continuous and differentiable. The solution of this static optimization problem defines the indirect utility of both parties in period t as a function of the debt at the beginning and at the end of the period. for a given fiscal deficit the consumer faces no uncertainty. For expositional purposes.10 In Section 1 of the Appendix several useful results regarding the partial derivatives of c*. x*. g* andf* are established. x*(b" b'+l). ) and h( .+l)' This function is identical for both parties..

Since generically ct '" 0. even if he cannot make binding commitments. 0. the absence of commitments does not matter: Proposition 1. (13) (14) e V~(b.e.:o Ail¥i and Ai are arbitrary weights such of public expenditures is a function of and a (18) of a. that I. Proof With an infinite horizon. For a discussion of this issue see Persson. ) and R~( .(b.) with respect to its ith argument. b.408 REVIEW OF ECONOMIC STUDIES Suppose that a policymaker. = b'+1 = b) we obtain: R.. II II We now turn to the problem of choosing the optimal composition ditures. with a zero initial government debt. In equilibrium a social planner balances the budget in every period.+I) + 8Ve(b. It is easy to show that Proposition 1 applies identically to this case for any choice 11. where Rf(·).+1 {Re(b" b. Outside of the steady state.+I)}. 2. Hence.. In this case the optimal policy with commitments would always balance the budget. Substituting the expressions equation (15) reduces to: b) =0. If bo'" 0. If the social planner knows with certainty the distribution consumers he maximizes: of public expenof the l¥ i across T~oo.+I) with respect to its first argument. From (13) and (14) it follows that in the steady state (i. Consider the steady state first. where N is the number of consumers.) The first-order conditions are: R. The optimal composition satisfies the following condition: (17) a = I. is certain of being reappointed each period. (15) for R. as in Lucas-Stokey (1983).+I) + 8V~(b'+I) = = = the social planner faces the following problem of (12) Maxb. for b. denotes the partial derivative of R (. if bo'" 0 the absence of commitment matters. then Proposition 1 needs to be rephrased so as to apply only to the steady state. We now show that. say party D. ) derived in section 2 of the Appendix. dynamic programming: Ve(b.:I Ai = 1. b)+8R~(b.) R~(b" b'+I). i = 1. (16) where ct is the partial derivative of c*( b. as well as in the finite horizon case. equation (16) can be satisfied if and only if b = 0. the planner runs a surplus (if bo> 0) or a deficit (if bo < 0) until the condition of zero outstanding debt is reached.( . . and not just in the steady state.Tabellini (1990).(b" b. A slight generalization of these arguments shows that the same results hold period-by-period.

Thus. and by the second-order conditions: sign g~} = sign {(R~(bt. increasing the probability of re-election raises the marginal cost of issuing debt (the right-hand side of (20». In more colourful terms. today's government can force its successor to "pay the bills" and spend less on the public good that is worth nothing to today's government.Ri". bl is a strictly decreasing function office in period 0 is reappointed.}. the consumers face no uncertainty about TI. note that the cost of leaving debt to the future consists of two components: the future tax distortions associated with the higher taxes and the reduced public consumption. we restrict each government to select strategies contingent only on the .A in the Appendix establishes that gf < O.2) in the Appendix. (21) where the second equality follows from (A.0)]. of the probability that the party in (19) Proof Suppose D is in office in period 0 and recall that P is the probability that D wins the elections. Thus.0)+(1-P)RN(bt. O).1. II Intuitively.) and R N ( • ) defined in Section 3 are independent of P. The right-hand side is the expected marginal cost of inheriting debt tomorrow. O»} = sign {hggf(bt. Lemma l. with gf being the derivative of g* with respect to its first argument. The Two-Period Case In the last period (period 1) both parties collect the same tax revenue.b. Suppose that party D holds office at the beginning of period o. (bt. abl/ aP < O. Equation (20) implicitly defines the equilibrium level of debt. discounted to the present by 8.1) and (A. By the implicit function theorem applied to (2). Hence. The proof for when party R is in office in period zero is symmetric. since they inherit the same debt and have to leave zero end-of-period debt. bl. in period 0. It follows that the interest rate and the functions Re(. the larger the probability of re-election.ALESINA & TABELLINI 5. as a function of P: Proposition 2. 0) . The amount of debt that this party chooses to leave to period 1 (bl) can be found by solving the following problem (recall that bo = 0): Max . 5. by leaving debt to the future. The left-hand side of (20) is the marginal utility in period 0 of leaving debt to the future.2.)+8[PRe(bt. the more the party in office internalizes the cost of leaving debt to the future. The second component is born only if the party currently in offce is reappointed next period and can choose the desired public good. V(0)=Re(0. and the smaller is the equilibrium debt. ALTERNATING DEFICITS AND DEBTS 409 GOVERNMENTS 5. The first-order condition can be written as: (20) The second-order conditions are assumed to be satisfied. To see why. The Infinite-Horizon Case In order to avoid the multiplicity of equilibria that inevitably arises in infinite-horizon dynamic games.

This implies that we do not explore reputational equilibria. we can ask what are the consequences on public debt of changing the probability of electoral outcomes. the political equilibrium described in Section 6 would not apply to this interpretation of P.1 of Lemma l. one can reinterpret P as being the (exogenous) probability of re-election for the current administration. Alesina (1987. + 8(PV~b + (1. The equilibrium that we consider is analogous to the notion of Markov-perfect equilibrium defined by Maskin. they do not fully internalize the costs of leaving debt to their successors.) R N tb. the argument of Proposition 2 still applies. Thus: VN (b. Thus. and not just in a neighbourhood of P =!. the tax rate and the level of public expenditure is the same for both parties. irrespective of which party it belongs to. the right-hand side is the expected marginal cost of inheriting debt tomorrow. In this case. 13.A in Section 1 of the Appendix is satisfied. In this case the optimization problem faced by the two parties is identical. but presumably not on. (23) The first-order conditions are: (24) Equation (24) has exactly the same interpretation of (20): the left-hand side of (24) represents the marginal utility of leaving debt to the future. However.l" From (24) we obtain: Proposition 3. . the Pareto frontier of the game. Proof See Section 3 of the Appendix. The condition alluded to in the text of Proposition 3 is needed to ensure that the total derivative of debt on public expenditure is negative in the steady state.+. As in the previous section.) are independent of p. alternating governments which disagree over the composition of public consumption issue more public debt than the social planner.P) VN (b'+I». Thus. Alternatively.) and RN (. The answer is that.+I) + 8PVe(b. In the two-period model. the deficit. As in the two-period case.410 REVIEW OF ECONOMIC STUDIES stock of debt outstanding at the beginning of its term of office. the results of this section would hold for any value of P. these costs correspond to the payment of interest on the stock of debt. Here.) = where Ve(. b'+I) + 8(PVe (b'+I) + (1. these costs take the form of higher taxes and lower public consumption in order to repay the debt in the final period.) = Maxb. 14.) are value functions of the party if elected and if non-elected respectively. 1988b) and Ferejohn (1986) consider reputational equilibria in repeated static political games. The intuition is still as in the previous section: since governments are not certain of reappointment. the steady-state level of public debt is always positive and locally stable if the sufficient condition c. in a neighbourhood of P = ~ under the same condition of Proposition 3 plus an additional weak sufficient 12.13 The dynamic programming problem solved by the party in office is: Ve(b.P) VN (bt+l)]. It can be conjectured that such equilibria would bring the economy closer to. (22) and VN ( .+I) + 8(1. Thus.P) Vbb) < O.{Re(b" b. The second-order condition now requires: R~.V We can characterize the solution only in the neighbourhood of P =~.Tirole (1988) (but see footnote 6). as in the two-period model. this eliminates consumers' uncertainty about future taxes so that Re(. the only difference is about which public good is supplied. In a neighbourhood of P =~.

A democracy in which citizens disagree about the composition of public expenditures exhibits higher deficits and debt accumulation than an economy with a social planner who is appointed for ever.. = L:~o8'[u(c. This argument is developed more in detail in Alesina (1988b). 0. bl)] = h(g*(O. under certain conditions. and suppose that elections are held at the beginning of period 1. Since voter i votes for party D if and only if the expected utility with D elected is not lower than if R is elected. Electoral uncertainty arises because of the uncertainty about who is the median voter. both parties can only credibly announce to the voters the time-consistent fiscal policy characterized in the preceding sections. provided that there is some uncertainty about the preference of future median voters. to derive P endogenously.~' Let am be the value of a corresponding to the median voter: party D wins if and only if am . bl))(2aj -1). and the horizon is finite. Uncertainty about "'m may be due.aJh(f*(O. The proof is available from the authors..a )h(g.) + v(x. The same results can be shown to apply in the infinite-horizon case. P is not a function of bl•15 Note that there is no convergence of the parties towards the same policy. for instance. bl)) - (1.. as in Ledyard (1984). 6. thus consumption and leisure are the same in period 1 regardless of the electoral outcome. . (25) holds if and only if aj. that imply a random voting turnout. we have that voter i votes D if and only if: (without loss of generality we assume that indifferent voters vote for party D): [ajh(g*(O. 7. even in a direct democracy. (25) Since h(g*( bl)) > 0.. though with opposite weights. given a probability distribution for am..)]. It is shown that a bias towards deficit survives.a)h(f.~. Thus: WD WR L:~o8'[u(c. the stock of public debt issued by either party in the steady state is a decreasing function of the probability of that party winning the elections. in particular. Both parties choose the same tax rate in period 1.) + ah(g. Tabellini-Alesina (1990) study a related political model in which the median voter theorem applies.. to individual specific shocks to the costs of voting.) + v(x.ALESINA & TABELLINI DEFICITS AND DEBTS 411 condition. POLITICAL EQUILIBRIUM In this section we analyze voting behaviour.) + (1. Consider first the two-period case. Specifically. Since binding commitments are not available.) + ah(j. EXTENSIONS In this section we discuss two extensions: (i) We generalize the objective functions of the two parties to the case in which they care about both public goods. We can summarize the results of sections 5 and 6 in the following: Proposition 4. = (27) (28) 15.)].) + (1. the probability that party D will be elected is given by: (26) From (26) it is apparent that P is a constant from the point of view of period zero.

this strategic interaction generates a sub-optimal path of government debt. by contrast. From a normative point of view. Recent empirical results by Roubini and Sachs (1989a. They show that the "conservative" policymaker (i. the one which likes less public expenditure) may choose to leave deficits in order to force its "liberal" successor to spend less. Since a larger value of a means that the two parties are more polarized._ k> o. In our paper. Thus we impose that f?. the larger is the deficit bias." In both cases.e. we consider! < a < 1 (the alternative case is completely symmetric). They consider the case of two policymakers with different views about the level rather than the composition of government expenditure. between the two parties the . CONCLUSIONS Public debt can be used by a policymaker to influence the choices of its successors. like the power function. This paper does not provide an argument in favour of balanced budget 16. different countries' experiences can be related to differences in the degree of political polarization. the "liberal" policymaker may leave a surplus to its conservative successors. b) on OECD countries are generally supportive of this line of research. the disagreement concerns the composition of public spending. This condition is stronger than decreasing absolute risk aversion. In particular. This tendency is stronger the greater is the degree of political polarization and of downward rigidity in public spending. the government tends to overissue public debt relative to the case of full agreement. these results contain some suggestions for institutional reforms._ k> 0 and g?. both parties run a surplus. in the political stability. These constraints may reflect institutional or technological factors limiting the flexibility of the government decision process. Related results have been independently obtained in a recent paper by PerssonSvensson (1989). 17. If this condition is not satisfied. More detail about that role played by this condition in a related model can be found in Tabellini-Alesina (1990). this paper shows that fiscal deficits are the aggregate outcome of the political conflict between different groups of citizens. Conversely. 8. (ii) We assume that a minimum level k> 0 of both public goods must be provided. If there is disagreement between alternating governments. More generally. but it is satisfied by a large class of commonly used utility functions. According to our results. (2) If neither of the constraints is binding a deficit bias still emerges if the following condition on the function h(·) applies: A(g) == -hgg/(hg)2 is decreasing in g. From a positive point of view. Our results can be summarized as follows (details are available): (1) If the constraints on g and f are binding in the sense that each party supplies only the minimum amount k of the good it likes the least. the results reported above can contribute to explaining why different countries pursue very different debt policies under similar economic conditions. If a> 1 the results obtained with a = I are strengthened since party D(R) attributes negative utility to good f(g): thus neither party ever would supply a positive amount of this good (an analogous argument holds for a <0). and in the flexibility of the government decision process concerning public consumption. and it gives rise to a deficit bias irrespective of which party is in office.16 To fix ideas.412 REVIEW OF ECONOMIC STUDIES for any 1 > a > 0. the results of the previous sections generalize directly. The greater is the degree of polarization larger is the deficit bias. it can be shown that the larger is a. we have: Proposition 5.

APPENDIX 1. or a recession). A practical normative implication is that different political parties should agree on a set of contingencies that would allow them to deviate from a balanced budget when in office (such as the occurrence of a war./ab" c~=acllab'+I' Gc=aG/ac" and so on. Let ct=ac. b)+d(b. II of Re(. To prove that the steady-state debt is positive. However. namely that fiscal deficits (surpluses) can be used to smooth tax distortions optimally over time. x* and g* are continuously differentiable. In this respect. we do not add anything to the existing literature on optimal fiscal policy. with P =! and using (22) and (23) in the text. On this issue.3) .2) Finally. (hg-uJ =---/j[u H c c (c '+1 ) (A. Finally use condition (C. these normative prescriptions lead once again to the well known policy dilemma of choosing between simple rules and discretion. II once the expressions 3. and P: b'+1 = Bib.+.) follows from the fact that the maximization is performed on a compact feasible set. Continuity (A. Strict concavity can be proved along the lines of Stokey-Lucas-Prescott (1989)-details are available upon request.A. Each administration should be allowed to incur deficits only if the prespecified contingencies materialize.l).) Differentiability follows from the fact that c*. and from the continuity of u(·). then: (i) gt < 0 and xt > 0 (ii) ct>O and gt(b. Details of the algebra are available from the authors.1) --=Re=---H ab 2 1+1 s R" (hg-uc) H c b. Lemma 1. (C. by definition of R N. Proof of Proposition 3. aR N / ab = ucct + v.xt . note that. we obtain: (A. P). From the envelope theorem and from the first-order conditions from which equation (10) in the text is derived we obtain: xn.ALESINA & TABELLINI DEFICITS AND DEBTS 413 amendments. v(·) and h(·). which can be shown to be positive for ct and xt have been substituted in it. the relevant set of contingencies to be included in the escape clauses can be very hard to identify and serious problems of enforcement and monitoring may arise. the second-order conditions of footnote 9 and the condition in footnote 10 referring to an upward-sloping labour supply function to sign all these partial derivatives. The proof of local stability (i. the paper does support the view that each government should not have complete discretion in its choice of how much deficit to leave to its successors. Needless to say. that IBII < l ). 2. Proof of Lemma 1.1) b)<Ofor b~O if: I~< I Proof First apply the implicit I Hxx(hg-Uc)+VxxHc+UccHAhg-Vx)/Ucl Hx-Hc theorem to the following . We will use the following notation: BI = aB(b" p)/ab" and so on. as implied by the application of the implicit function theorem to (10). 2. reproduced from the text: (9) (10) function two equations Recall that by the resource constraint: g1 = -(c1 + i = 1. is available from the authors. Equation (24) in the text implicitly defines b'+1 as a function of b.e. (See the theorem of the maximum in Hildenbrand (1974).

TABELLINI.(b. (1979). G. TABELLINI. "Why a stubborn conservative would run a deficit: Policy with time-inconsistent preferences". R. BARRO. (1977) Democracy in Deficit (New York: Academic Press). "Principles of fiscal and monetary policy". (1986). J. "U. (Cambridge: Cambridge University Press).J. Mass. (A. and MELTZER. and an editor for useful comments on preliminary versions of this paper. ALESINA. Susan Vitka. in NBER Macroeconomic Annual 1988" (Cambridge. "Macroeconomic Policy in a Two-Party System as a Repeated Game". (1980).+[) = R~(b.5» also rules out the possibility of a negative debt in the steady state. G. BRENNAN. 796-806. LUCAS. (ed. and ALESINA. BARRO. L. A. and TABELLINI. Journal of Monetary Economics. A slightly more elaborate argument (which again makes use of equation (A. (1987) Macroeconomics. A. J. (2nd ed.2) one obtains R~= -aR~. N. Journal of Political Economy. 199-220. 713-33. II Acknowledgement. "Credibility and policy convergence in a two-party system with rational voters". 193-222. Substituting this expression in (A. . 104.5) By Lemma LA. 325-346.+" and using (A.4) and using (11) of the text again to simplify. HILDENBRAND. L. (1989).5) is satisfied if and only if P = 1. "A theory of dynamic oligopoly I: Overview and quantity competition with large fixed costs". American Economic Review.: Harvard University Press).S. unless P = 1. 365-74. R. 55-93. MASKIN. Needless to say. R.. Quarterly Journal of Economics. L. and SVENSSON. G. 101. In the steady state. BARRO. G. Public Choice. An earlier version of this paper was circulated as NBER Working Paper No. 14. 37-49. W. capital flight and political risk".) High Public Debt: The Italian Experience. "The end of large public debts". Quarterly Journal of Economics. (Cambridge: Cambridge University Press). BUCHANAN. (1989) Recursive Methods in Economic Dynamics (Cambridge. "A political theory of government debt and deficits in a neo-Ricardian framework". we obtain: V. and Hb. CUKIERMAN. John Van Huyck. 549-569. Econometrica. R. 117-134. Journal of Monetary Economics. ALESINA. 56. American Economic Review. American Economic Review. prices and budget deficits in the United Kingdom. REFERENCES ALES INA.) (New York: John Wiley and Sons). J. and TIROLE. (1984). Journal of International Economics. "The politics of intergenerational redistribution". LUCAS. (1988a). (1989). (1985). Daniel Heyman. 2308. A. PERSSON. (1988c). (A. "On the determination of the public debt". (1988b). A. by the envelope theorem. E. 651-678. and WAGNER. T. G. 78. 88.+[. and PRESCOTT. "Optimal fiscal and monetary policy in an economy without capital". A. (1983).5) into (23) of the text and use (AA) Suppose that the steady-state level of public debt is O. R. two referees. John Lott. ALESINA. (1989). FEREJOHN. b. A. PERSSON. 80. BARRO. PERSSON. 27. (1986). 1730-1918" (University of Rochester Working Paper). "Time-consistent fiscal policy and government cash-flow". 17. in Giavazzi-Spaventa. The power to tax.: MIT Press). we obtain: (A. "Macroeconomics and politics". (1986). and STOKEY. (1989). E. deficits since World War I". R. "Government spending. 940-947. "Incumbent performance and electoral control". A. Journal of Monetary Economics. and BUCHANAN. Herschel Grossman. Mass. (1987). and TABELLINI. LUCAS.5) then yields a contradiction. N. Journal of Political Economy (forthcoming). "Voting on the budget deficit. 79. J. Computing Hb. "External debt. 50. N. Credibility and Politics (London: Harwood). 12. Since B[ < 1. STOKEY.: Princeton University Press). 5-26. (1974) Core and Equilibrium of a Large Economy (Princeton.. 87. gt + g! < 0 and gt < O.414 REVIEW OF ECONOMIC STUDIES Moreover. (1990) Macroeconomic Policy. Scandinavian Journal of Economics. interest rates. We are grateful to Richard Cantor. A. (1990). T. (11). R.e. ALESINA. in the case of the social planner.+2)' Substitute (A. and SVENSSON. R. i. (1988). T.

Sign up to vote on this title

UsefulNot usefulAre you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.