SENGAMALA THAYAAR EDUCATIONAL TRUST WOMEN’S COLLEGE

SUNDARAKKOTTAI, MANNARGUDI - 614016.

(Accredited with A grade by NAAC)
(An ISO 9001 : 2015 Certified Institution)

PG & RESEARCH DEPARTMENT OF ECONOMICS

CLASS :III B.A ECONOMICS

SUBJECT NAME : ECONOMICS OF GROWTH AND DEVELOPMENT

STAFF NAME : K.ARCHANA

Module-II : Technological Progress

Technological progress embodied and disembodied technical progress
What is Technological Progress?
Technological progress refers to the discovery of new and improved methods of producing goods.
Changes in technology lead to an increase in productivity of labor, capital, and other factors of production.
Technology refers to the process through which inputs are transformed into outputs.
A technological change involves the invention of technologies and their release as open
source via research and development, the continual improvement of the technologies, and the diffusion of the
technologies throughout the industry or society.
Phases of Technological Progress
1. Invention
Invention is the act of creating new technology. It involves a new scientific or technical idea, and the
means of its embodiment or accomplishment. To be patentable, an invention must be novel and have utility.

2. Innovation
Innovation may be used synonymously with “invention” or may refer to discovering a new way in which
to use or apply existing technology. Everett Rogers thought of innovation as an idea, behavior, or product that
appears new to its potential adopter. There are five main attributes of innovative technology: Relative
Advantage, Compatibility, Complexity, Trialability, and Observability.
Relative advantage means the product or behavior is perceived as being better than the alternatives by the
person adopting the innovation. Better can mean a lot of different things. It can be a device that can peel a
potato faster so it saves time or a seat belt that offers the advantage of greater safety.
Compatibility refers to how the innovation aligns with the adopter’s lifestyle.
Complexity is how easy or difficult in innovation is to understand. The easier an innovation is to understand
and use, the more likely it is to be adopted. Complex innovations face an additional challenge to mainstream
adoption.
1
Trialability refers to the process of testing the innovation to see if, or how well, it works. Extensive testing
usually occurs before an innovation is adopted or taken to market.
Observability involves seeing the product or behavior in action. It can demonstrate how it can be used. It is
easier to get potential adopters to simply observe an expensive product like a car than it is to get all of them in
one for a test drive. Also, the more people around you that you see using a product, the more likely you feel
like buying that product too.
3. Diffusion
Diffusion pertains to the spread of technology throughout a society or industry. It is the process by which a
new idea, product, or behavior is accepted by the market. Technology diffusion means the spread of
usage/application of new technology from its current user to others. The diffusion of innovation theory,
introduced by Everett Rogers, explains how different groups of people adopt innovation in different ways, in
order to best suit their own needs or desires.
How to Measure Technological Progress
One of the most common methods used to measure technological progress is through the Solow Residual. The
Solow Residual method works under the assumption that all changes in output that can’t be explained by
changes in the capital stock or changes in the number of workers must be due to technological progress. The
method uses a simple linear regression to estimate growth.
1. Regress output on capital and labor using simple linear regression.
2. The regression residuals are TFP growth. (Total Factor Productivity – the ratio of aggregate output (e.g.,
GDP) to aggregate inputs)
Technical progress can be classified into two parts:

• Embodied Technical Progress: improved technology which is exploited by investing in new equipment.
New technical changes made are embodied in the equipment.
• Disembodied Technical Progress: improved technology which allows increase in the output produced
from given inputs without investing in new equipment.
In the real world, many innovations do not require replacing the entire or some part of the equipment. It
can be improved for better use depending upon the change required. Hence technological progress, embodied
or disembodied, is matter of degree.
Technological Progress and Economic Growth
Technological change is the most important factor that determine rate of economic growth. It plays a
important role than the capital formation. It is the technological change which can bring about continued
increase in output per head of the population. Thus it is the prime-mover of economic growth.

Technological change or progress refers to the discovery of the new and improved methods of
producing goods. Sometimes technological advances result in the increase in available supplies of natural
resources. But more generally technological changes result in increasing the productivity of labour, capital and
other resources. The productivity of combined inputs of all factors is called total factor productivity. Thus
technological progress means increase in total factor productivity. As a result of technological advance, it
becomes possible to produce more output with same resources or the same amount of product with less
resource.

2
 But the question arises as to how the technological progress takes place. The technological progress
takes place through inventions and innovations. The word invention is used for the new scientific discoveries,
whereas the innovations are said to take place only when the new scientific discoveries are commercially used
for actual production of goods. Some inventions may not be economically profitable to be used for actual
production.

It follows from above that technological change brings about an increase in output per head. Thus
technological change, or more precisely technological progress, is the change in the production process which
results in an increased output per unit of labour. Technological change causes a shift in the production function
embodying all known techniques. Technological change must be distinguished from a change in technique.
While by technological change we mean advance in knowledge resulting in improved methods of production,
change in technique refers to the use of a different but already known method of production.

The process of economic growth involves the increase in the production of goods and services. Increase
in production can be achieved either through the use of more resources and/or through the realization of higher
productivity by means of using the resources of labour, capital and land more efficiently. Technological change
helps to promote growth in both these ways. It can help in the discovery of new natural resources in the
country and thereby enhances the productive potential of the country. Technological change also increases the
productivity of available resources.

For instance, it can find out the productive uses of land that hitherto has been regarded as infertile or it
can discover new economic use of a raw material that had previously been considered as useless. But, as
explained above, technological change more generally results in higher productivity of resources.
Technological change raises the productivity of worker through the provision of better machines, better
methods and superior skills.

Table 8.1 gives the percentage increase in labour productivity in a number of countries during the
period 1970-1989. By bringing about increase in productivity of resources the progress in technology makes it
possible to produce more output with the same resources or the same amount of output with less resource.
Progress in technology causes improvement in technology through the provision of better machines, better
methods and enhanced skills.

It is technology which underlies the process of producing new things with the existing resources or
using the existing resources in new ways. This is what Schumpeter means when he says that, “the slow and
continuous increase in the national supply of productive means and savings is obviously an important factor in
explaining the course of economic history through the centuries, but it is completely overshadowed by the fact
that development consists primarily in employing existing resources in a different way, in doing new things
with them, irrespective of whether those resources increase or not. It is important to emphasize that newly
discovered techniques lead to the increase in output per worker.”

Productivity of workers depends upon the quantity and quality of capital tools with which they work.
For higher productivity the instruments of production have to be technologically efficient and superior. The
technological options open to an economy determine the input mix of production. A commodity can be
produced by various technologies. The quantity and quality of capital, skills and other factors required for
production is directly dependent on the efficiency of the technique of production being used. Also, the
managerial and organisational expertise has to be in line with the technological requirements of production.
Thus, technology in the present stage of economic development is an indispensable factor of production.

3
 This is the age of technology. The developing countries are obsessed by the desire to make rapid
progress in technology so as to catch up with the present-day developed countries. Frantic efforts are being
made to install improved technology in agriculture, industries, health, sanitation and education; in fact in all
walks of human life. Indeed, the newly emerging nations have come to regard technology as a bastion of
national autonomy and as a status symbol in the international community.

It will be noticed that during 1970-89 rate of increase in labour productivity for the United States is less
than other countries except Canada and Sweden. Slow growth in productivity in the Unites States during the
period (1970-89) had been a major concern and was the subject of debate among the economists. It will be
further observed that during this period Japan had the highest growth rate in labour productivity of 6%
followed by Italy and Netherlands.

Denison in his empirical study has estimated the sources of increase in real GNP between 1929 and
1982 of the United States. He found that technological change contributed 28% to the average annual growth
in real GNP which was 2.9 per cent over this period. The estimates of contribution of various factors such as
labour, education and capital are presented in Table 8.2.

From the table it will be seen that increase in the quantity of labour contributed to 32 per cent to 2.9 per
cent annual growth in the United States. The other Factors contribute to the annual growth by raising
productivity of labour.

Technical progress manifests itself in the change in production function. So a simple measure of the
technical progress would be the comparison of the position of production function at two points of time. The
technological change operates upon the production function through improvements of various sorts such as
superior equipment, an improved material, and superior organisational efficiency. Also, the technological
progress may express itself in making available new products. The availability of a new product in many cases

4
could be construed as a change in the production function as it can be regarded equivalent to a more efficient
way of meeting an old want.

Technological Change and Capital Formation:

The process of technological progress is inseparably linked with the process of capital formation. In
fact, both go hand in hand. Technological progress is virtually impossible without prior capital formation. It is
because the introduction of superior or more efficient techniques requires the building up of new capital
equipment which incorporates new technology. In other words, new and superior technology can contribute to
national product and its growth if it is first embodied in the new capital equipment.

The capital accumulation has, therefore, been called the vehicle for the steady introduction of new
technology into the economy –

It may be noted that Adam Smith viewed technological progress as a rise in productivity of workers as
a result of increase in division of labour and specialization. This leads to the growth in national income. But it
was J.A. Schumpeter who laid great stress on the role of technological innovations in bringing about economic
growth. He visualised technical innovations in bringing about economic progress. It is the entrepreneur who
carries out the innovations and organises the production structure more efficiently. As, according to
Schumpter, the innovations occur in spurts rather than in a smooth flow, economic progress is not an
uninterrupted process. The pace of economic progress is punctuated by the pace of innovations. Since
innovations are governed by a host of factors, therefore, it is difficult to predict about the economic progress
resulting from technological changes.

Prof. Rostow has proposed four stages in the development of an economy.

These stages are:

(i) Traditional society;

(ii) Preconditions for takeoff;

(iii) Drive to maturity and

(iv) Stages of high mass consumption.

It may be noted that the economic transformation of the society from one stage to another involves,
along with other things, a change in the level and character of technology. In the present age of greater
specialisation it is the technology factor that underlines all major aspects of the modern productive apparatus
such as decision making, production programming, and skill requirements market strategy.

The role of technological progress in economic development has not only been large but also increasing
over time. It is beyond any doubt that the inventions and changes in production techniques played more
important role in the late eighteenth and later centuries than in the earlier ones. The technical innovations
provided the motive force for the industrial revolution in Western Europe. While the innovations mainly arose
because of input shortages, they tended to overcome the impediments to mass production and the transport
bottlenecks in the movement of industrial raw materials and labour. There were also improvements in
managerial organisation with a view to reduce costs of production. The adjustments in the product designs also
took place to meet the variations in the wants of the people.

5
 Technical progress is the prime mover of economic growth. A continuing technical progress ensures a
continuing growth. Where there is no technical progress, capital accumulation alone may fail to bring growth
beyond a point. For capital accumulation without technical progress can bring about growth in income per
head only so long as the available labour force has not been fully equipped with the latest techniques available,
and only so long as all the wants of the consumers have not been supplied to the full extent. Once that point is
reached where all labour has been equipped with the latest techniques and all wants of consumers have been
met to the full extent, growth would come to an end unless technical progress occurs.

Technological Progress to Overcome Stationary State:

It is now widely accepted that technological change raises productivity and that a continuous
occurrence of technological change will enable the economy to escape from being driven to the stationary state
or economic stagnation. Classical economists like David Ricardo and J.S. Mill expressed fear that the increase
in the stock of capital will sooner or later land the economy into stationary state beyond which economic
growth will come to an end. Classical economists remained occupied with the idea of a stationary state because
they did not take into account technological progress that could postpone the occurrence of a stationary state
and ensure continued economic growth. Indeed, if technological progress continuously takes place, demon of
stationary state can be put off indefinitely.

We can diagrammatically show how the technological change will raise productivity and thereby
suspend the occurrence of stationary state with the help of Hicks’s representation of classical model. Let us
first represent the classical model of stationary state. Let us consider Fig. 8.1 where along the X-axis stock of
capital is measured and up the Y-axis rate of return on capital is measured. This rate of return on capital is in
effect of the marginal productivity of capital. It may be noted that rate of return is here taken to be quite akin to
the ‘ rate of interest’ which was regarded by the classical economists as the mechanism which determines
investment to be made in any period of time.

Now, let us suppose that the economy has, at a particular moment, a stock of capital equal to K1, in Fig.
8.1. Given this stock of capital, the rate of return which the businessmen would be able to earn from a new
investment would be somewhat less than r3 .However, if the economy is to be kept growing through capital
accumulation while maintaining full employment, the rate of interest must be less than r3; let us say it is at the
level r2. In this situation, it would be profitable for the businessmen, in the given period, say, one year, to
undertake investment (through borrowing or with their own funds) to an extent that results in a net addition of
K1 K2 to the capital stock. Therefore, at the end of the year, the economy would come to have K 2 amount of
capital stock. It is possible now to draw a curve CC in Fig. 8.1 which indicates the different rates of return
obtainable from the different levels of capital stock.

6
 Assuming that the supply of labour is constant, in the absence of technological progress and
consequently rate of return curve CC remaining the same, the expansion in the stock of capital from K1 to
K2 will bring down the rate of return on capital from r3 to r2. As a result, the rate of return on new investment
will now be slightly below r2 after K1K2 investment has been undertaken in the year. Now, if there are still net
positive savings in the community which is likely the case when capital stock has not yet sufficiently
increased, then the rate of interest will fall below the level of r2. With the fall in the rate of interest below r2 and
the rate of return on new investment being nearly equal to r2 (i.e., only slightly less than r2), it will be
worthwhile to undertake new investment in the next year and in this way to maintain full employment in the
next year.
According to the classical economists, rate of interest will continue falling gradually over time till it
reaches a level at which both net saving and net investment have fallen to zero. In Fig. 8.1 this will take place
when the capital stock has increased to K3 and rate of interest has fallen to r1. Since there are no more net
savings and investment, the further expansion in the stock of capital and therefore the process of economic
growth will cease to occur and the economy will reach what classical economists called stationary state.
However, according to the classical economists, the stationary state is a distant end which may not be reached
at all. This view of ‘classicals’, according to Hicks, is based on the idea that the curve CC is highly elastic, so
that the distances such as K1 K2 are, indeed, very small. Thus, capital accumulation takes place by very small
steps and can go on for a very long time before the rate of return finally falls to r1.

Stationary State can be postponed through Technological Progress:

It may be noted that the businessmen would continue undertaking investments so long as the capital
stock is less than K3 .For, until K3 level of capital stock is reached, the rate of return is high enough to induce
businessmen to use the net savings for undertaking new investments. However, once the K3 level of capital
stock is reached, net savings would have fallen to the zero level and consequently both net investment and
capital accumulation will cease to occur. While in this situation, net investment would necessarily have come
to an end, gross investment may continue. But the gross investment does not add to capital stock; it is made
just for the purposes of replacements of worn-out capital. When the stationary state has been reached, net
investment would be zero because net savings are zero. Also, the population, income and standards of living
would stabilise at a constant level.

Now, the importance of technological progress is that it can suspend the occurrence of stationary state
equilibrium. Classical economists underestimated the role of technological progress in preventing the
occurrence of stationary state equilibrium. When technical progress occurs it will raise the productivity of
capital and labour. As a result of the rise in productivity, the rate of return curve CC (i.e., marginal productivity
curve of capital) will shift upward.

This implies that it will be profitable to accumulate more capital and thereby suspending the occurrence
of the stationary state. This is illustrated in Fig. 8.2 where to begin with economy is in stationary state
equilibrium at point S1 with rate of return on capital equal to r. Suppose that a new technological advance takes
place which is embodied in a new machinery with which output per unit of labour is greater than before. As a
result of this technological advance, the marginal productivity curve (i.e., curve of rate of return on capital)
will shift upward.

7
 As will be seen from Fig. 8.2, with the shift in the rate of return curve from C1C1 to C2C2 the rate of
return on capital has risen to r1. With the rate of return on new investment being nearly equal to r1 and the rate
of interest being equal to r, it will be profitable to undertake new investment. As a result, stock of capital will
increase to the level K2 and the whole of the given labour force has been equipped with new machines
corresponding to the new technology.

With the increase in stock of capital to K2, the rate of return will fall to r. Thus when the rate of return
again falls to r level, the net saving and the net investment will be reduced to zero and a new stationary state
equilibrium will be reached at S2. With the stock of capital equal to K2, suppose there occurs a new burst of
technical progress which causes the CC curve to shift further upward to C 3C3. This will cause the capital
accumulation to take place further to K3 .Thus a result of technological progress and more capital
accumulation, output, income and the standards of living of the people will rise.

It follows from above that technological progress holds the key to economic growth. As technological
progress takes place, capital accumulation will proceed further and economic growth will take place. Without
technical progress, capital accumulation can raise the output and income to a limited extent. With constant
labour force and the given technology, capital accumulation can lead to the increase in output so long as all the
workers of the labour force are not equipped with the latest capital equipment embodying the latest technology.
After this, technological progress can lead to the growth of output and income.

Thus, Professors Stonier and Hague rightly state – “Technical progress, therefore, turns out to be even
more essential to economic growth than capital accumulation. Where there is no technical progress, capital
accumulation can lead to growth in income per head only so long as labour force has not been fully equipped
in accordance with the latest techniques. Once that happens, growth comes to an end. It is technical progress
alone which ensures that the stationary state is not merely very far off but will never arrive so long as technical
progress continues.”

Technological Change – Disembodied and Embodied:

The factors which cause sustained growth of national income and per capita income is an important
issue that has been the concern of economists. Capital accumulation, growth in labour force and technological
progress contribute to the growth of an economy. American economist R.M. Solow used the aggregate
production function that relates the level of output to the levels of various inputs such as growth of capital,
labour and technological progress over time. Solow used the following production function-
Y = A (t) F (K, L)…… (1)

8
 Where, Y stands for output, K for capital and L for labour. The term A (t) represents technological
change which is taken to depend merely on time, that is, as time passes, A (t) increases showing that over time
technological progress enables more to be produced with the given quantities of capital and labour. With such
specifications of production function, technological progress does not cause any change in the relative
marginal productivities of capital and labour, that is, such technological progress brings about equal increases
in the productivities of the two factors and is said to be Hicks’ neutral in the sense that it favours neither capital
nor labour.

But the important thing to note in the production function given above is that it represents technological
progress which is exogenous and disembodied. By exogenous technological progress we mean it comes from
outside the model and, therefore, the term A (t) is shown outside the production function. Disembodied
technological change raises the productivity of both old and new machines (i.e., capital goods) and does not
depend on the rate of new investment.

The disembodied technological change is purely organisational in nature which permits more output to
be produced from the given inputs without any new investment. That is, disembodied technological change
causes a shift in the production function without disturbing capital-labour ratio. It is through organisational
changes that disembodied technological change enables more output to be produced with existing inputs being
used more effectively.

Disembodied technological change is represented in Fig. 8.3 where on the horizontal axis capital-labour
ratio (K/L), that is, capital per worker is measured and along the vertical axis output per worker(Y/L) is
measured .In the beginning, production function curve Y = f1(k/L) is given and, according to this, with a capital
per worker equal to K/L, level of output Y1 is being produced.

Now, suppose that technological progress takes place and production function curve shifts upward to
OY’ that is, to Y’ = f 2(K/L). According to the new production function curve Y’ = f2(K/L), with the given
capital-labour ratio, K/L more output Y2 is produced. Similarly, with any other capital-labour ratio on the new
production function curve OY’, i.e., Y’ =f2(K/L), more output will be produced as compared to that on the
production function curve OY.

9
Embodied Technological Change:

In the embodied technological change, the increase in investment or capital accumulation is regarded as
the vehicle of technological progress. Technological progress increases the productivity of new machines built
in any period compared with old machines built in the previous periods. This implies that embodied
technological change does not cause increase in the productivity of machine already in existence.

About embodied technological change, Hahn and Matthews in their review of the theory of economic
growth write, “Technological change is embodied in new machines. Machines unalterably embody the
technology of their date of construction. Machines built in different dates are therefore qualitatively dissimilar
and cannot in the general case be aggregated into a single measure of capital. A separate production function is
needed for each vintage. Total output is the sum of Output of all vintages in use”.

Production function of embodied technology can be written as –

Yt =F (Kt, Lt, At,) … (2)

Where, output Yt depends on the amount of capital (Kt), labour (Lt) and the level of technology (At).
Note that in the production function (2), the term At for technology appears inside the production function as
one of the endogenous inputs.

It is important to note that in the production function (2), the relationship between output and
technology differs from the relationship between output and other inputs, capital and labour. This can be better
understood by considering this production function in case of an individual firm which can be stated as –

Yit= F (Kit, Lit, At) … (3)

The production function (3) of the individual firm shows that output of an individual firm depends not
only on its own level of capital (Kit) and labour (Lit) but also on the economy-wide level of technology.
Expressing production function in this way, the technological progress in the economy will increase the
productivity of all firms including the firm whose production function (3) is given above.

In this formulation of production function, the technology is assumed to progress endogenously instead
of being an exogenous change. Thus, in the concept of embodied technology, production function depends on
the investment in new capital. Technological progress fosters inventions and improvements in machines. Thus
new improved technology is embodied in investment in the new machines.

Neutral, Capital– Saving and Labour– Saving Technological Change:

It is important to know the difference between labour–saving and capital–saving technological change.
But before explaining labour-saving and capital-saving technological change, it will be useful to make clear the
meaning of neutral technological changes, since it is neutral technology which is the dividing line between
labour-saving and capital-saving technological change. In economics neutral technological change (or
innovation) has been defined in two ways, the first by J.R. Hicks and the second by R.F. Harrod.

These two concepts of neutral technological change are explained below:

Hicks’ Neutral Technological Change:

Hicks defined neutral technological change in the context of static price theory, especially in
connection with the theory of wages. According to Hicks, neutral technological change (or neutral innovation)
is one where if certain ratio of two factors, say, labour and capital, is being used to produce a given output, the
effect of neutral technological change is to bring about increase in the marginal productivity of each factor in
10
the same proportion. Hicksian concept of neutral technological change (neutral innovation) has been shown in
Fig. 8.4 where along the X-axis, capital per man is measured and up the X-axis output per man is measured.

To the left of the origin O along the X-axis we represent the ratio of marginal products of labour and
capital such as OR. The curve OQ1 depicts the production function (total product curve) according to some
existing technology. Suppose a technological change takes place and as a result production function shifts
upward to OQ2.

With production function OQ1 and, before technological change in Fig. 8.4 the equilibrium is at point T
on the production function OQ1 where wage rate is equal to OW which is equal to the marginal product of
labour. It should be noted that in Fig. 8.4 the slope of the curve TR represents the marginal product of capital
as along the X-axis, capital per man is measured.
Now, the slope of the tangent, TR = TK/RK = OW/OR

It follows therefore that the marginal product of capital = OW/OR

If ν stands for marginal product of capital, then

ν = OW/OR

Manipulating the above equation we have –

OR = OW/ ν

As stated above, OW represents marginal product of labour and ν the marginal product of capital–

Therefore,

OR = Marginal Product of labour /Marginal Product of capital

It is, therefore evident that the distance OR measures the ratio between the marginal product of labour
and the marginal product of capital in the position of equilibrium.

11
Now, if technological progress is to be Hicks-neutral, then technological change causing shift in production
function from OQ1 to OQ2 should leave the ratio between the marginal product of labour and marginal product
of capital unchanged at any given value of the capital-labour ratio. Since at the equilibrium position T in Fig.
8.4 the capital-labour ratio (i.e., capital per man) is OK, Hicks’ neutral technology requires that at the capital-
labour ratio OK, the ratio of marginal product of capital to marginal product of labour remains the same (i.e.,
equal to OR) when technological change has taken place and production function has shifted to OQ2.
This implies that when the new production function OQ2 and at point ‘T’ corresponding to capital-labour ratio
OK, if a tangent is drawn, it should pass through the same point R. It will be seen that such is the case in Fig.
8.4. With the slopes of the tangents RT’ and RT to production functions Q2 and Q1 being equal to OR implies
that technological change brings about change in the marginal products of labour in the same proportion. It will
also be observed from Fig. 8.4 that as a result of the technological progress and shift in production function,
the output per man has increased and also the equilibrium wage rate has increased from OW to OW ‘. Thus
both labour productivity and capital productivity has increased as a result of technological progress.

Capital-Saving and Labour-Saving Technological Change:

From Hicks’ definition of neutral technological change, the concepts of capital-saving and labour-
saving technical change can easily be understood. A capital-saving technological change (capital-saving
innovation) is that which, at a given capital-labour ratio, raises the marginal productivity of labour relative to
the marginal productivity of capital. This means that it would now be possible to produce a given level of
output with less capital relative to labour as a result of technological progress.

Thus technological change is capital-saving in the Hicksian sense when the ratio– Marginal Product of
Labour/ Marginal Product of Capital or (ΔQ/ ΔL) / (ΔQ/ ΔK) increases, where ΔQ stands for a small change
output in response to a small change in labour input (ΔL) and a small change in capital input (ΔK). Thus
ΔQ/ΔL and ΔQ/ΔK will measure marginal product of labour and marginal product of capital respectively.

Note that capital-saving technological change generally implies labour-using technological change as
labour is substituted for capital.

In terms of Fig. 8.4 if capital-saving technological change occurs, the production function will shift in
such a way that the tangent drawn at point T ‘ (corresponding to the given capital-labour ratio OK) then that
tangent will pass through the left of point R indicating that the ratio of marginal product of labour to marginal
product of capital will increase. As a result, the given output can be produced with less capital relatively to
labour. Therefore, it shows capital-saving innovation (i.e., technological change).

Labour-Saving Technological Change:

According to Hicks, labour-saving technological change is one which reduces the marginal product of
labour relative to that of capital. In other words, labour- saving technological change (i.e., labour-saving
innovation) occurs when at a given capital-labour ratio,

The ratio, Marginal product of Labour/ Marginal product of Capital or (MPL/ MPK) declines.
This implies that in case of labour-using technological change it would now be possible to produce a
given level of output with less labour relative to capital following a change in technology. Note that labour-
saving technological change generally implies capital-using technological change. In other words, in this case
capital-intensity of production will increase. In terms of Fig. 8.4 the production function curve will shift in
such a way that the tangent drawn at the new production function corresponding to the given capital-labour
ratio OK will pass through the right of point R.

12
Harrod’s Neutral, Labour-Saving and Capital-Saving Technological Change:
It has been shown by R.F. Harrod, a prominent British economist, that Hicks’ concept of neutral change
is heavily dependent on factors such as elasticity of demand for factors and products which are quite unrelated
to the real character of technological change. Besides Hicks’ concept of neutral, technology has been
developed in the framework of static economic theory. Therefore, an alternative concept of neutral
technological change which is extensively used in the theory of economic growth has been put forward by R.F.
Harrod in his now well-known work “Towards a Dynamic Economics”.

While Hicks’ concept of neutral technological change considers the ratio between the marginal
products of the two factors when capital-labour ratio remains constant, Harrod states his concept of neutral
technological change in terms of the relationship between rate of profit and capital-output ratio. According to
Harrod, neutral technological change is that which leaves the capital-output ratio unchanged when rate of profit
(assumed to be equal to marginal product of capital) remains constant. Thus when there is technological
progress and rate of profit remains constant, technological change will be Harrod-neutral only if capital-output
ratio also remains unchanged.

It follows from the definition of Harrod’s neutral technical change that when rate of profit remains
constant after the technological change, technological progress would be labour-saving in the Harrodian sense
if capital-output ratio rises. This means that now same output can be obtained by using more capital. Given the
level of output, the use of more capital implies the substitution of capital for labour.

On the other hand, when there is technological progress and rate of profit remain the same, the
technological progress would be capital-saving in the Harrodian sense if capital-output ratio falls. This implies
that now the same output can be produced with less capital. Obviously, there is saving in capital as a result of
substitution of labour for it. It is worthwhile to note that since in Harrod’s concept of neutral technological
change, rate of profit and capital-output ratio remain constant, it means that distribution of income between
profits and wages will remain unchanged following such a type of technological change. In terms of Harrod’s
growth model, the implication of capital-output ratio remaining constant when a neutral technological change
occurs is that relative shares of wages and profits in national income remains constant as balanced growth of
national income takes place.

Harrod’s concept of neutral technological change is graphically illustrated in Fig. 8.5. In this figure,
capital per man is measured along the X-axis and output per man is measured up the Y-axis. Initially, the
production function curve is OP1 showing the change in total output as capital per man increases. Suppose the
economy is in equilibrium at point T1 on the production function curve OP1. At point T1 on the production
function curve OP1, capital-output ratio is OK1/OY1.

The straight line OR passes through the origin. Therefore, all points on the straight line OR such as
T1 and T2 represent the same capital-output ratio. Thus capital-output ratio at T2 is equal to OK2/OY2.

Hence,

OK2/OY2=OK1/OY1

Suppose there is technical progress and as a result production function shifts from OP1 to OP2. The
economy will now be in equilibrium at some point, say T2, on the new production function OP2. As the points
T1 and T2 lie on the same straight line OR passing through the origin, capital-output ratio at the equilibrium
position before and after the assumed technical change has remained unchanged.

13
 But for technological change to be neutral in the Harrodian sense, not only capital-output ratio should
remain the same but also rate of profit should remain unchanged as production function changes and we move
along the ray OR representing a given level of capital-output ratio. Now, the rate of profit which is assumed to
be equal to the marginal product of capital will be given by the slopes of the production function at the relevant
points.

If the technological change which has raised the production function from OP1 to OP2 is to be Harrod-
neutral, then slope at point T1 on production function OP1 and slope at point T2 on production function
OP2 must be the same. The slope of functions at T1 and T2 would be the same if the tangents drawn at these
points are parallel to each other. This is exactly the case in Fig. 8.5 where tangent drawn at point. T 1 on the
production function OP1 before technical change and tangent drawn at point T2 on production function
OP2 after the technical change are parallel.

Nature of Technological Change and Income Distribution:

In Harrod’s growth model, neutrality of technological progress implies that income distribution
between wages and profits will remain constant as equilibrium growth of the economy occurs. In Harrod’s
growth model, there are two factors of production-labour and capital – and growth with neutral technological
progress causes increase in incomes of both workers and capitalists. Income of workers increase as rate of
wages per man hour rises due to increase in their productivity as a result of technological change.

Profits of capitalists increase because they accumulate more capital assets as growth proceeds, though
rate of profit per rupee worth of capital remains the same. Thus, commenting on Harrod’s growth model with
neutral technological progress Stonier and Hague write, “The workers’ standard of living will rise because his
wage rate has risen. The capitalist’s standard of living will rise too but it will do so because he gets much the
same rate of return on each pound’s worth of a bigger stock of capital.”

However, if in Harrod’s growth theory capital-saving technological progress occurs, the share of profits
in national income will fall and that of wages will rise. On the other hand, if labour-saving technological
progress occurs the share of profits in national income will rise and share of wages will decline during the
process of growth.

14
Hicks Harrod, learning by doing production function approach to the economic growth
Models of Technical Change in Economic Growth
Technical progress plays an imperative role in influencing the pace of economic growth. It is the
technical change which results in an increased output per unit labour. It signifies a comprehensive phenomenon
and, therefore, denote different things in different contexts.

Technical change refers to change in the production function embodying all techniques.

Thus, technical change in the context of economic development must result in more output for the same
resources or the same amount of output. It could occur following a change in any of the production variables.
Thus, it may result due to a change in kinds of physical capital, in the quality of labour or even in the
organisation of these resources.

Harrod-Dommar model is based on the assumption of fixed coefficients of production and gives rise to
the knife edge problem. Kendrick, Kaldor and Solow and others have been the most consistent critics of this
approach who have tried to demonstrate the role of technological changes in the growth of an economy.

In this way, the nature of the technical progress is most important determinant of the individual factor
productivities. It brings a proportionate increase in the productivities of all the factors. Therefore, the model of
technical change is based on the controversy over neutral and non-neutral technical change.

Neutral and Non-Neutral Technical Changes:

A technical change is said to be neutral if it is neither capital saving nor labour saving i.e. it is neutral in
its effect in the sense that neither of the two factors become more or less important at the margin. There are
two definitions of neutrality. One is given by Prof. Hicks and the other by Prof. Harrod.

A. Hicks views on Neutrality. According to Prof. Hicks, neutrality is “An invention which raises the
marginal productivity of labour and capital in same proportion”. Thus, a technical change is neutral if the
ratio of marginal product of capital to that of labour remains unchanged at constant capital labour ratio.
A technical change is termed as labour saving if it raises the marginal product of capital relative to that of
labour at a constant capital labour ratio.

15
Hick’s neutral technical change is explained in the figure 1 by comparing points on two different;
production functions:

The output per man q is represented along with vertical axis and capital labour ratio is represented
along with horizontal axis. OQ measures the marginal product of labour and capital. OP is the production
function before technical change and OP1 is the production function after technical change. Taking, production
function OP, the slope of tangent QTD measures the marginal product of capital and OT measures the marginal
product of labour.

Since slope QT shows the marginal product of capital, say u, we can express it as:

u = OT/ OT or OQ = OT/u

Hence OQ measures the ratio between marginal product of labour OT and marginal product of capital
‘u’, Hicks neutral technical progress requires that if technical change shifts the production function upwards
from OP to OP1, the ratios of two marginal products must be same on vertical line from X-axis like KE where
it passes through the production functions at points D and E, respectively.

Hicks Neutral Technical Progress:

The condition is that the tangent QE on the higher production function OP, must originate from the
point Q to left of O, like the tangent before the technical change. In the figure, the tangent QE on the
production function OP1originates from Q.

When both the tangents QE and QD on production functions OP and OP1originate from Q, only then
the ratios between marginal product of labour and capital will be equal i.e. the ratio between the marginal
product of labour and capital after the technical progress (OW1/u1) must be equal to the ratio between the
marginal products of labour and capital before technical progress (OW/u)Therefore, the ratio between the
marginal product of labour and capital is equal at points D and E on the vertical line KE.

This situation can be expressed as under:

Q = Q (t) f (K, L)

where Q — Total output

K — Inputs of capital

L — Inputs of labour

A (t) — Index of technical progress. It measures accumulated effects of shift overtime and is an increasing
function of t.

On the basis of the definition of Hicks neutrality, we can define labour saving and capital saving technical
changes, which Mrs. Joan Robinson’s terms biased technical progress.

Capital—Saving Technical Change:

A technical change is capital saving if it raises the marginal product of labour relatively to capital, at
constant capital labour ratio. The given output will require less capital relative to labour.

16
This is illustrated with the help of a diagram 2 given below:
This shows that a shift in production function resulting from the capital saving technique would be one where,
for a given K

This type of technical change is capital saving, i.e. it would be possible to produce a given level of
output with less capital relative to labour than before the introduction of technique.

If the amount of capital used is reduced absolutely and that of labour rises, the technical change is
absolutely capital saving and labour using. On the other hand, if the technique leads to a fall in the amounts of
both the inputs, but the fall in the input of labour is less compared to that of capital, the technique is said to be
relatively capital saving.

Labour-Saving Technical Change:

A technical change is labour saving if it raises the marginal product of capital relative to labour at
constant capital labour ratio. The given output would require less labour relatively to capital i.e. for a given K
and it would be possible when the technical progress causes a rise in the productivity of capital proportionately
more than that y of labour i.e. it is possible to produce a given level of output with less labour relative to
capital. The figure 3 given below shows the case of labour saving technique.

Where t is the isoquant before the technical progress and t1 after the technical progress. If the amount of
labour used is reduced absolutely and that of capital rises than the technical change will be absolutely labour
17
saving and capital using. On the other side, if the technique leads to a fall in both factors but the fall in amount
of labour is proportionately more than that of capital, the technique is said to be relatively labour saving.
The precise manner in which relative and absolute amounts of labour and capital used will change as a result of
technical change will depend upon the factor elasticities of substitution and product elasticities of demand.

Therefore, it will depend upon the elasticities of substitution between capital and labour in the
economy, for these will help to determine what effects the technical change has on the prices of the two
factors. It will also depend upon the elasticities of demand for the product of various industries that make up
the economy.

In Hicks Neutral Technical progress, the factor shares remain constant if factor proportions and relative
remunerators of labour and capital are constant. If he consider figure 1, it implies that between D and E if the
slope of the production function OP1 at E is greater than the slope of production function OP at D in the same
proportion as the output KE is greater than KD, then the technical progress is Hicks Neutral.
This means that when the amount of capital is changed, the marginal product of capital increases in the same
proportion as total output. In other words, between D and E, the proportion of total output which is paid out of
profits and wages remains constant. Again, it is also as when the elasticity of substitution between labour and
capital is equal to unity.

Criticism:
Hicks neutrality has been criticized on many grounds.

Firstly, it is a rigid type of definition even when large factors of production are involved. Secondly, the
dependence of Hicks Neutrality on demand elasticity and substitution elasticities makes it a cumbersome tool
of analysis. It is quite unrelated to the intrinsic character of the innovation itself such as the elasticity of
demand for product and factor.

Lastly, the Hicksian neutrality is built within the frame work of static economic analysis.

Harrod’s View on Neutrality:

Prof. Harrod in his book “Towards a Dynamic Economics” has defined neutral technical progress in a
different way. His definition is based on capital output ratio. According to him, neutral technical progress is
one which leaves capital output ratio unchanged, provided that rate of profit remains constant. Thus, Harrod’s
neutrality of technical progress requires the constancy of both the rate of profit r and capital output ratio K/Y.
If the rate of profit remains unchanged and the capital output ratio increases, the technical progress would be
labour-saving. On the other hand, if after the technical progress, the capital output ratio falls, while the rate of
profit remains constant, the technical progress would be capital saving.

18
Harrod Neutrality is explained with the help of a diagram 4.

In this diagram, the capital per man (k) is measured along X-axis and output per man (q) is along Y-
axis. OP is the production function before the technical change and OP is after technical change. The capital
output ratio at D on production function OP is OK1/OY1 and at point E on production function OP’ is
OK2/OY2.

Since the ray OL passes through both points D and E, the capital output ratios are equal i.e. OK1/OY1=
OK2/OY2. Harrod’s neutrality also requires that the rate of profit must remain constant along with a constant
capital output ratio after technical progress.

This implies that the marginal productivity of capital must be same at both points of D and E on
production functions OP and OP’, respectively which, in turn, requires that the slope of the production function
OP at point D must be equal to the slope of the production function OP’ at point E. In other words, the tangents
at D and E must be parallel to each other.

In figure 4, the tangent MN at D is parallel to the tangent M’N’ at point E i.e. the marginal products of
capital at D and E are the same, Thus, Harrod neutral technical change as shown by shifting the production
function OP upwards to OP’ depicts the equality of the capital output ratio at D and E as represented by the ray
OL passing through them.

Thus, the equality of slopes at D and E shows that the rate of profit is constant.

Harrod’s definition of neutral technical progress is superior to that of Hicks because it is applicable to a
dynamic situation rather than a static situation. It forms a most significant part of the theory of economic
growth because it uses the concept of capital output ratio which is indispensable in modern growth analysis.

In Harrod neutral technical change, there is no direct reference to labour as it is entirely based on the
relationship between capital and output but capital labour ratio and output labour ratio may change without
technical change, Therefore, in Harrod’s neutrality, the rise in output per machine would be in the same
proportion as the rise in the output per man.

Harrod’s neutrality implies that the distribution of income between profits and wages does not change
because under perfect competition, the factors are paid according to their marginal product. As it leaver the
marginal product of capital unchanged, the factor shares as between capital and labour, must also remain
unaltered.

The assumption of constant capital output ratio implies that the capital stock and labour force must
grow at same rate. It follows that the incomes of the capitalists would grow at same speed at which the wages
of the workers rise. If a technical progress is capital saving in Harrod’s sense, this will raise the share of labour
in the national output and reduce that of capitalists at a constant rate of interest.

On the other hand, a labour saving technique will reduce the share of labour in national output and
increase that of capitalists with constant rate of interest.

19
Harrod’s neutrality can be shown in the form of production function as:

Q = [F K, A (t) L]

Here Q is a function of F and K and A (t) L means that the given constant return to scale on equal
proportionate rise in capital (K) and in effective labour units [A (t) L] must lead to proportionate rise in
national output Q. The constant rate of interest increases the efficiency of labour in whole economy.

However, “With population growth, there is an increase in the number of men at work, Harrod
neutral technical progress, the GNP rises at given rate.” The difference is that with Harrod’s neutral
technical progress, income per head increases, with population growth it remains the same.
As Mrs. Joan Robinson and Uzawa have shown on the strict definition, Harrod’s neutral technical progress
raises income at same rate whatever may be the level of the capital output ratio. It is this which measures
technical progress. Solow has shown that the Harrod’s neutrality can be purely capital augmenting technical
progress with production function.

Q = F [A (t) K, L]

A (t) — Index of technical progress.

The technical progress is both Hicksian Neutral and Harrod’s Neutral.

If the elasticity of substitution between labour and capital is unity and there is no change in the
distribution of income, it is Harrod’s Neutral. The neutral change is in Hick’s sense if the given labour force
capital remains unchanged and distribution of income is the same. It is Harrod-neutral if with given labour
force, capital increases in the same proportion as national output and distribution of income is the same.

Disembodied Technical Change:

During 1956, M. Abramovitz wrote papers followed by J.W. Kendrick and R.M. Solow to measure the
contribution of technical change to economic growth. They treated technical change as disembodied.

Disembodied technical change is purely organizational which permits more output to be produced from
unchanged input, without any new investment. It refers to any kind of shift in the production function that
leaves balance between capital and labour undisturbed in the long run.

The production function for such technical change is written as:

Q = F (K, L; t)

Q — Output

L — Labour input

K — Capital input

t — Technical change.

20
Taking Hicks neutral technical change as a basis, Solow postulated the production function in special
form as:

Q = A (t) F (K, L)

(A) — Index of technical change.

Thus, accordingly, “Such a production function implies that technical progress is organizational in
the sense that its effect on productivity does not require any change in the quantity of inputs. Existing
inputs are improved or used more effectively”.
The growth rate of output is equal to the rate of technical change plus a weighted average of growth rate of
capital and growth rate of labour Assuming linear homogenous production function, these weights add to one
and we have Q/Q =A/A + at (K/K) + (1- at ) L/L

Where does indicate time derivatives and ‘at’ is the capital elasticity of output.

Solow proceeded to focus on rate of technical change. He stated “By using data on the share of capital
and labour and the rates of growth of capital per head and output per head, the contribution of the residual is
obtained after calculating the contribution of capital.

This residual is attributed to technical progress”. He concluded that during 1940-49 the average growth
rate of output per head in United States could be attributed 12.5 per cent to increase in capital per worker and
the residual 87.5 per cent to technical change.

Criticism:

The above conclusions tried to undermine the rate of investment in contrast to technical change in growth
process.

According to Phelps, “The results of this approach provided a wave of investment pessimism”.
Whereas Rosenberg, writes, “They provide a wide response on the part of the economists wakened as it were,
from their dogmatic slumber.” They become sceptical about such a large size of the residual. Abramowitz
admitted, “It is a measure of out ignorance”.

Griliches puts that, “Residual approach is not of much use in understanding the growth process as it is
based on the concept of a production function which is not very useful if it is not a stable production function
and if there are very large unexplained shifts in it”. Critics further observed that, “Residual approach tended to
ignore other influences like improvements in the quality of labour due to education”.

This approach is based on the unrealistic assumptions of perfect completion, constant returns to scale
and complete homogeneity of capital stock. Denison, Kendrick, Griliches tried to quantify and break down the
residual into further components.

They contended that the residual was not a catch all and that changes in output were due to changes in
the quantities and qualities of inputs, in economies of scale and advances in knowledge rather than the result of
technical change, assuming a stable production function.

21
Embodied Technical Change:

Solow modified the residual approach himself based on disembodied technical change where in capital
stock is regarded as homogeneous and technical change floats down from the outside.

According to F.H. Hahn and R.C.O. Mathews, “In this model, new capital accumulation is regarded as
the vehicle of technical progress. Technical progress increases the productivity of machines built in any period
compared with machines built in previous period, but it does not increase productivity of machines already in
existence. Technical progress is embodied in new machines. Machines alterably embody the technology of
their date of construction. Machines built at different dates are therefore, qualitatively dissimilar and cannot in
the general case be aggregated into single measure of capital. A separate production function is needed for
each vintage. Total output is the sum of output of all the vintages in use”.

Assumptions:

This model is based on the following assumptions:

1. Capital stock consists of machines of different vintages or built at different dates.

2. New machines are more productive than old ones.

3. Technical change proceeds at some given proportional rate.

4. Machines embody all the latest knowledge at the time of construction but do not share in any
subsequent improvements in technology.

5. Technical change effects only new machines.

6. Only gross investment in new machines is considered in the model and the production function is linear
homogeneous.

7. Technical change proceeds at some given proportional rate.

8. The production function is linear homogeneous of the Cobb- Douglas type.

The total output Qv (t) at times t from the machines of each vintage v is given by Cobb-Douglas production
function

22
 “The picture is one of the continuum of capital goods of various vintage and corresponding
productivity, subject to an exponential life table”, according to Solow. At each moment of time the labour
force is reshuffled over the existing capital goods. This total output is determined by integrating over all layers
of capital stock”.

Solow states that if we assume competition in the labour market, all homogeneous labour must receive
the same wage regardless of the age of capital on which it operates.

23
 Solow calls J as an effective stock of capital which is productivity weighted sum of all surviving capital goods
representing all early technological levels.

But the capital goods of smaller vintages receive a smaller weight than new capital goods. He further
adds that average age of capital can be lowered by increasing the saving rate and thereby average quality of
machines in use can be raised. Thus output per man can be raised.

Appraisal:

The model of disembodies technical change is based on the assumption that capital stock is completely
homogeneous. New machines are better than old machines and technological progress is embodied in new
machines. In the former, capital labour ratios change at all times along the Cobb-Douglas production function.

But in the latter, once a machine is constructed, it has labour requirements.

In other words, “Each machine is designed to be worked with given crew of men and the size of
crew cannot be changed”.
Mrs. Joanson has rightly observed that this model is said to have ex-ante substitutability between labour and
capital and ex-post fixed coefficients or no ex-post substitutability. Phelps puts that it is a putty day model-
putty ex-dante and clay ex-post. On the other hand, the model of disembodies technical change is one of ex-
ante and ex-post substitutability or a partly model.

Limitations:

The drawbacks of technical change model are listed as below:

1. It is based on the assumption of perfect competition and hence fails to consider factor market
imperfections.

2. Solow assumes that machines depreciate exponentially. But Stieglitz points out that this may be
reasonable assumption for telephone poles but not for machines.

3. It does not take into account the influence of wage expectations on machine construction. An investor
forms expectations of wage rates extending into future before constructing a machine. In that case, the
real wage rate will not equal the marginal productivity of labour on the machine of a given vintage and
type, but it will equal the average output per man on the least efficient machine.

4. The entire model is based on the hypothesis that machines are of different types and new machines are
better than old ones.

5. The assumption on which this model is based relate to fixed labour requirements. This is unrealistic for
an economy with a higher output per man which may have lower capital labour ratio.

6. It concentrates only on technological progress embodied in new machines and ignores the problems of
inducing innovations through the process of learning and investments in research.

24
TOTAL FACTOR PRODUCTIVITY AND GROWTH ACCOUNTING
Introduction
Colonial Americans were very poor by today’s standard of poverty. At the time of American
Revolution, Gross Domestic Product per capita in the United States was $765 approx. (in 1992 dollars). During
the next two centuries, income increased hysterically, pushed upward by the Industrial Revolution, and Gross
Domestic Product per capita had increased to $26,847 by 1997. But GDP did not grow smoothly (see fig. 1.1
and table 1.1), but it has been consistent at an average 1.7 % annual growth rate. Moreover, the transformation
that was brought by Industrial Revolution made Americans leave the farm and take up manufacturing jobs and
even more jobs in the economy’s service sectors.
One of the basic aim in economics research is to comprehend this revolution. Response has been given
by theorists in the form of various models. Marxian and neoclassical growth theories allot higher weightage to
improvement in productivities drifted by technological advancement and the organization of production. On
another side, New Growth Theory and another branch of economics of neo-classics—capital and investment
theory—affix foremost significance to the increase in investments in human capital, fixed capital and
knowledge. The dichotomy between technology and capital formation carries over to empirical growth
analysis. Talking on general terms, economist had two significant jobs: one, huge task of constructing inputs &
outputs’ historical data; and two, to measure the degree to which output growth is, in fact, due to technological
factors (“productivity”) versus capital formation. This last undertaking is sometimes called “sources of growth
analysis” and is the intellectual framework of the TFP residual, which is the organizing concept of this survey
Total Factor Productivity
Total Factor Productivity (TFP) is the share of output not explained by the quantum of Inputs used in
the process of production. Therefore, how efficiently and effectively inputs are used in the production process
determines its level. In economics, total-factor productivity (TFP), also called multi-factor productivity, is a
variable accounting for effects in the output that is not due to commonly used inputs like capital and labour. If
we take into account all inputs, then TFP is the representation of economy’s long-term technological change or
technological dynamism. We can’t measure TFP in a direct manner. But Solow residual, a residual, is
responsible for measuring the effects in total output which inputs do not cause. The equation below (in Cobb-
Douglas form) represents total output (Y) as a function of total-factor productivity (A), capital input (K), labor
input (L), and the two inputs' respective shares of output (α and β are the capital input share of contribution for
K and L respectively). If either of A, K or L is increased, output will also get increased. Labor and capital input
are tangible, TFP seems to be more intangible ranging from technology to worker skill that is, human capital.

𝑌 = 𝐴 × 𝐾 𝛼 × 𝐿𝛽
Growth in technology and efficiency are termed as two of the main sub-categories of TFP, the former
owns "special" features which are inherent like positive externalities and nontrivialness which improves its
position as an economic growth driver. Total Factor Productivity is often seen as the real driver of growth
within an economy and studies reveal that whilst labour and investment are important contributors,
Total Factor Productivity may account for up to 60% of growth within economies. TFP is more
accurately measured in long term, since TFP can vary substantially from one year to another. It has been
shown that there is a historical correlation between TFP and energy conversion efficiency
TFP growth is usually measured by the Solow residual. Let gY denote the growth rate of aggregate
output, gK the growth rate of aggregate capital, gL the growth rate of aggregate labor and alpha the capital
share. The Solow residual is then defined as gY − α gK − (1 − α) gL. The Solow residual accurately measures
TFP growth if (i) the production function is neoclassical, (ii) there is perfect competition in factor markets, and
25
(iii) the growth rates of the inputs are measured accurately. TFP plays a critical role on economic fluctuations,
economic growth and cross-country per capita income differences. At business cycle frequencies, TFP is
strongly correlated without put and hours worked. The TFP model produces an explanation of economic
growth based solely on the production function and the marginal productivity conditions. Thus, it is not a
theory of economic growth because it does not explain how variables on the right-hand side of the production
function— labor, capital, and technology—evolve over time. However, Solow himself provided an account of
this evolution in a separate and slightly earlier paper. He assumed that labor and technology were exogenous
factors determined outside the model, and that investment is a constant fraction of output. Then, if technical
change is entirely labor augmenting and the production function is well-behaved, the economy converges to a
steady-state growth path along which both output per worker and capital per worker grow at the rate of
technical change. Cass (1965) and Koopmans (1965) arrive at essentially the same conclusion using different
assumptions about the savinginvestment process. Both of these “neoclassical” growth models produce a very
different conclusion from that of the TFP model about the importance of technical change in economic growth.
In the neoclassical growth models, capital formation explains none of the long-run, steady-state growth in
output because capital is itself endogenous and driven by technical change: Technical innovation causes output
to increase, which increases investment, which thereby induces an expansion in the stock of capital. This
induced capital accumulation is the direct result of TFP growth and, in steady state growth, all capital
accumulation and output growth are due to TFP. While real-world economies rarely meet the conditions for
steady-state growth, the induced-accumulation effect is present outside steady-state conditions whenever the
output effects of TFP growth generate a stream of new investment.
Measurement of TFP
The residual is a valid measure of the shift in the production function under the Solow assumptions.
However, because the TFP residual model treats all capital formation as a wholly exogenous explanatory
factor, it tends to overstate the role of capital and understate the role of innovation in the growth process. Since
some part of the observed rate of capital accumulation is a TFP-induced effect, it should be counted along with
TFP in any assessment of the impact of innovation on economic growth. Only the fraction of capital
accumulation arising from the underlying propensity to invest at a constant rate of TFP growth should be
scored as capital’s independent contribution to output growth.
The distinction between the size of the residual on the one hand and its impact on growth on the other
has been generally ignored in the productivity literature. This oversight has come back to haunt the debate over
“assimilation versus accumulation” as the driving force in economic development. A number of comparative
growth studies have found that the great success of the East Asian Tigers was driven mainly by the increase in
capital and labor rather than by TFP growth .With diminishing marginal returns to capital, the dominant role of
capital implies that the East Asian Miracle is not sustainable and must ultimately wind down. However, these
conclusions do not take into account the induced capital accumulation effect. The role played by TFP growth is
actually larger, and the saving/investment effect is proportionately smaller.
Neoclassical growth models assume that innovation is an exogenous process, with the implication that
investments in R&D have no systematic and predictable effect on output growth. But, can it really be true that
the huge amount of R&D investment made in recent years was undertaken without any expectation of gain? A
more plausible approach is to abandon the assumption that the innovation is exogenous to the economic system
and to recognize that some part of innovation is, in fact, a form of capital accumulation. This is precisely the
view incorporated in the endogenous growth theory of Romar and Lucas. The concept of capital is expanded to
include knowledge and human capital and is added to conventional fixed capital, thus arriving at total capital.
Increments of knowledge are put on an equal footing with all other forms of investment, and therefore the rate
of innovation is endogenous to the model. The key point of endogenous growth theory is not, however, that
R&D and human capital are important determinants of output growth. What is new in endogenous growth
theory is the assumption that the marginal product of capital is constant—not diminishing as in the neoclassical
26
theories? It is the diminishing marginal returns to capital that bring about convergence to steady-state growth
in the neoclassical theory; and, conversely, it is constant marginal returns that cause the induced-accumulation
effect on capital to go on ad infinitum.
Most of the TFP studies that have incorporated product-oriented innovation into the residual have
focused on capital-embodied technical change. Nelson (1964) expressed the residual as a function of the rate of
embodiment and the average age of the capital stock. Domar (1963) and Jorgenson (1966) observed that capital
is both an input and an output of the production process, and the failure to measure capital in efficiency units
causes two types of measurement error: one associated with the mis-measurement of capital input and one with
the mis-measurement of investment good output. Surprisingly, the two errors exactly cancel in Golden Rule
steady-state growth, leaving the residual unbiased. Once it is recognized that product quality adjustments allow
consumer welfare parameters to creep into the TFP residual, the boundary between the supply-side conception
of the residual and the demand-side interpretation is blurred. If welfare considerations are permitted inside one
region of the supply-side boundary (and they must be, if the quality dimension of output is to make sense),
perhaps they should be permitted in other boundary areas, such as the netversus-gross output controversy,
where welfare arguments have also been made. After all, a high rate of real GDP growth, and hence a large
gross-output productivity residual, can be sustain din the short run by depleting irreproducible resources at the
expense of longrun welfare. Net output solves this problem by controlling for depreciation and environmental
damage; some believe that it thus provides a more accurate picture of sustainable long-run economic growth.
Does it not follow that a separate TFP residual based on net output is the appropriate indicator of the
contribution of costless technical innovation to sustainable growth? The short answer is “no.” Changes in
social welfare can be shown to depend on the standard gross-output concept of TFP, with no need to define a
net-output variant of TFP.
TFP residual can, in principle, be computed for every level of economic activity, from the plant floor to
the aggregate economy. These residuals are not independent of each other because, for example, the
productivity of a firm reflects the productivity of its component plants. Similarly, industry residuals are related
to those of the constituent firms, and productivity in the aggregate economy is determined at the industry level.
As a result, productivity at the aggregate level will increase if productivity in each constituent industry rises, or
if the market share of the high productivity industry increases (and so on, down the aggregation hierarchy). A
complete picture of the industrial dynamics of an economy would include a mutually consistent measure of the
TFP residuals at each level in the hierarchy and of the linkages used to connect levels
The task of constructing this hierarchy of residuals can be approached from the top down, in a process
that can be likened to unpeeling an order to reach lower layers of structure. Domar was the first to work out the
problem of “unpeeling” the TFP residual, and to recognize the complication introduced by the presence of
intermediate goods. This complication arises because plants and firms in each sub layer produce goods and
services that are used as inputs in the production processes of the plants and firms. As each layer is unpeeled,
the magnitude of these intermediate deliveries grows. For example, there are no intermediate goods in the
aggregate economy because there is only one industry at this level of aggregation, and all inter industry flows
cancel out. However, these inter industry flows “uncancel” in passing. However, these inter industry flows
“uncancel” in passing to the one digit industry level of detail. The iron ore delivered to the steel industry is
counted in the gross output of the extractive industries, and is counted again as part of the gross output of the
manufacturing industry. The sum of the one-digit industry gross output is therefore larger than total aggregate
output. The nature of this problem can be made more precise by observing that the total output of an industry
(plant, firm) is composed of deliveries to final demand plus deliveries of the industry’s output to the other
industries that use the good. On the input side, the firm uses not only labor and .capital, but also intermediate
goods purchased from other industries that use the good. Several studies have attempted to link the micro and
macro levels of analysis. Baily, Hulten, and Campbell (1992) used data from the LRD z into the weighted sum
of the plant-level residuals, it was found that the plants with rising TFP levels and plants with high preexisting
27
TFP levels were the main contributors to productivity growth. Firms with low preexisting TFP levels and
declining firms were dragon, productivity. The persistence of firms with both high and low levels of
productivity suggests a more complex view of industrial organization than the simple representative agent
model used to motivate aggregate TFP residual. The micro data also suggest a more complex productivity
dynamic in which the entry and exit of firms, as well as their expansion and contraction are important
dimensions.
The true character of the subject should be revealed. This is particularly important in the case of the
TFP residual, the true character of which has often been misunderstood by friends and critics alike. The portrait
painted in this paper reveals these essential features:
2. The TFP residual captures changes in the amount of output that can be produced by a given quantity
of inputs. Intuitively, it measures the shift in the production function.
3. Many factors may cause this shift: technical innovations, organization land institutional changes,
shifts in societal attitudes, fluctuations in demand, and changes in factor shares, omitted variables,
and measurement errors. The residual should not be equated with technical change, although it often
is.
3. To the extent that productivity is affected by innovation, it is the costless part of technical change that
it captures. This “manna from heaven” may reflect spillover externalities thrown off by research
projects, or it may simply reflect inspiration and ingenuity.
4. The residual is a nonparametric index number designed to estimate one parameter in the larger
structure of production, the efficiency shift parameter. It accomplishes this by using prices to estimate
marginal products.
5. The various factors comprising TFP are not measured directly but are lumped together as a “left-
over” factor (hence the name “residual”).They cannot be sorted out within the pure TFP framework,
and this is the source of the famous epithet, “a measure of our ignorance.”
6. The Divisia index must be path independent to be unique. The discrete-time counterpart of the
Divisia index, the Tornqvist approximation, is an exact index number if the underlying production
function has the translog form. The problem of path dependence is one of uniqueness, and this is not
the same thing as measurement bias.
7. The conditions for path independence are (a) the existence of an underlying production function and
(b) marginal productivity pricing. Neither constant returns to scale nor Hicksian neutrality are
absolutely necessary conditions, although they are usually assumed for convenience of measurement.
8. When the various assumptions are met, the residual is a valid measure of the shift in the production
function. However, it generally understates the importance of productivity change in stimulating the
growth of output because the shift in the function generally induces further movements along the
function as capital increases
9. The residual is a measure of the shift in the supply-side constraint on welfare improvement, but it is
not intended as a direct measure of this improvement. To confuse the two is to confuse the constraint
with the objective function.
10. When full equilibrium does not pertain, as in the midst of any lagged adjustment process, the
marginal equivalencies needed for successful aggregation do not obtain and it is likely that some of
the increases in productivity of labour and capital will be recorded as increases in the quantities of
labour and capital inputs.
It seems to us that, whatever TFP does measure and there is cause for concern as to how to answer that
question it emphatically does not measure all of technological change. In the long term, we are interested in
increases in output per unit of labour, resources (and waiting in the Austrian sense of the term). While people
are of course free to measure anything that seems interesting to them, the degree of confusion surrounding
TFP, particularly the assumption that low TFP numbers imply a low degree of technological dynamism, would
28
seem to us to justify dropping the measure completely from all discussions of long term economic growth.
Even if that does not happen, as we are sure it will not, every TFP measure should carry the caveat: there is no
reason to believe that changes in TFP in any way measure technological change.
Growth Accounting
Introduction
In economics, growth accounting is a procedure to measure the contribution of various factors to
economic growth. Growth accounting refers to the breaking down the rate of growth of total output of an
economy into contribution from the growth of such inputs as capital and labour and as well as technological
growth. Growth accounting is also related to the sources of growth. In an economy, the rate of technological
progress is indirectly compute, measured as a residual.
Robert Solow has introduced this methodology in 1957.Further the basics of growth accounting were
presented in Kendrick (1961), Denison (1962), and Jorgenson and Griliches (1967) to provide an overview of
this intellectual history, with stress on the development of the Solow residual. Hence, technological progress
plays a vital role in the economic growth of country.
Basically, growth accounting decomposes the growth rate of an economy's total output into two part,
firstly that which is due to increase in the amount of capital and labor and secondly, that which cannot be
accounted for by observable changes in factor utilization i. e. unexplained part of growth in total output is then
taken to represent increases in productivity or due to technological progress. It is also known Solow residual.
The residual is the difference between the growth rate of output and the measured growth rates of inputs.
In short,

Solow Residual or technical progress is the

residual between output growth rate and
weighted growth rates of inputs.

There are following three leading economists who have measured the contribution of the residual in terms
of sources of growth to the overall growth rate of the United States economy;

❖ Robert Solow’s Sources of Growth

❖ Denision’s Sources of Growth
❖ Jorgenson-Griliches Sources of Growth

In the present module, we will discuss Robert Solow’s sources of growth. Denision’s sources of growth
and Jorgenson-Griliches sources of growth will be discuss in the next module.

Assumptions of Growth Accounting

❖ It is based on constant returns to scale (Euler’s Theorem) i.e., Capital share + Labour Share = 1.
❖ It is based on perfect completion.
❖ Capital Stock is in complete homogeneity.
❖ The production function is linear and homogenous i. e. neoclassical production function.

29
 ❖ It breaks down the growth of output into the growth of the factors of production and technical change.
❖ This approach is based on long run period.
❖ Technical change is based on Hicks-neutral augmentation.

Growth Accounting or Solow Residual

R. Solow published a model in 1956 that representing a simplified but at the same time powerful
framework for the analysis of the causes and dynamics of economic growth. After one year, in 1957, he
published a paper entitled ‘Technological Change and the Aggregate Production Function’. This paper
indicates that growth of aggregate production is represented as a combination of the contributions of growth
rates of factors of production and technological change or total factor productivity. He separates variation in
output per head due to technical change from those due to changes in availability of capital per head.

R. Solow assumes technical change as disembodied, where capital as treats as homogeneous and he
also assumes that the technical change are exogenous. Disembodied technical change is capital augmented in
which existing capital is made more productive. Thus, the productivity depends upon the amount of capital
stock not on it age.

The production function for such technical change is written as;

𝑄 = 𝐹(𝐾, 𝐿,𝑡)

Q = Output

L = Labour Input

K = Capital

T = Technical Change

Taking Hick Neutral technical change as a basis, R. Solow postulated the following specification of
production function;

𝑌𝑡 = 𝐴𝑡 . 𝐹[𝐾𝑡 , 𝐿𝑡 ]------------------------------------------ (i)

Where, Yt = Aggregate production/ Total income

Kt = Stock of Physical Capital used in production

Lt = Amount of labour input

At = Level of technology

Equation (i) can be transformed as;

𝑌′𝑡 𝐴′ 𝒕 𝐾′ 𝑡 𝐿′
= = 𝑎𝑡 + 𝑏𝑡 𝐿 𝑡----------------------------------(ii)
𝑌𝑡 𝐴𝒕 𝐾𝑡 𝑡

The shares of capital and labour costs in total costs are at and bt respectively. So the of shares is equal
to one.
30
 The equation (ii) can be written as;

𝐴′ 𝑡 𝑌 ′ 𝑡 𝐾′𝑡 𝐿′ 𝑡
= − 𝑎𝑡 − 𝑏𝑡
𝐴𝑡 𝑌𝑡 𝐾𝑡 𝐿𝑡

Solow Residual / Technical Change

=
Rate of Growth in Aggregate Production
-
Growth Rate of Physical Capital and Labour output

Empirical Results

Robert Solow proceeded to focus on rate of technical change by using data on the share of capital and
labour and the rates of growth of capital per head. He stated that the contribution of the residual is obtained
after calculating the contribution of capital. The residual is attributed to technical progress. He concluded that
the average growth rate of output per head in United States could be attributed 12.5 per cent to increase in
capital per worker and the residual 87.5 per cent to technical change during the period 1940-49

Graphical Presentation

The latter calculation is also known as Solow residual. The diagrammatical representation of
technological change is given below. The third part of diagram indicates the output growth due to higher
productivity or Solow residual and due to increases in factor of production.

31
Examples of Solow Residual

For an example, consider Indian economy, whose total output grows at 5 per cent per annum.
Assuming that there are two factors of production i. e. capital and labour. During the same period its capital
stock grows at 7 per cent per annum and its labor force by 3 per cent per annum. The contribution of the
growth rate of capital to output is equal to that growth rate weighted by the share of capital in total output and
the contribution of labor is given by the growth rate of labor weighted by labor's share in income. The share of
capital's and labour in total output are 1/3 and 2/3 respectively. This means that the portion of growth in output
which is due to changes in factors is 0.07×(1 ⁄3)+0.03×(2 ⁄3) = 0.043 or 4.3 per cent per annum. This means
that there is still 0.7 per cent per annum of the growth in output that cannot be accounted for. This remainder/
residual are the increase in the productivity of factors that happened over the period, or the measure of
technological progress during this period.

Criticism of Solow Residual

Robert Solow has been criticized for his approach of measuring the residual/ technical change on the
following grounds;

❖ This approach is based on several unrealistic assumptions such as perfect competition, complete
homogeneity of capital stock and constant returns to scale.
❖ Solow estimates undermine the role of investment in context to technical change in process of growth.
According to Phelps, ‘The results of Solow approach produced a wave of investment pessimism.’
❖ Solow further admitted that there are index number problems involved in the measurement of every
variables in his measurement of the residual.
❖ This approach ignored the various components of sources of growth such as improvement in skill and
quality of labour force, improvement in technological level, change in composition of inputs and output
of an industry, investment in research and education and so on.
❖ Several economists point out that Solow emphasis the role of capital by assuming disembodied
technical progress, whereas the most significant advances in technical progress requires capital
embodiment. R. Solow himself admitted it in 1959.
❖ Abramovitz considered residual as ‘a measure of our ignorance’. While according to Rosenberg,
residual ‘provided a wide response on the part of economists wakened, as it were from their dogmatic
slumber.’
❖ Griliches pointed out that the residual approach is not of much use in understanding the growth process
because it is based on the concept of unstable production function which is the causes of very large
unexplained shifts in it.

Denison’s Sources of Growth

Introduction

Robert Solow (1957), Denison (1962), and Jorgenson-Griliches (1967) are the leading economists of
growth accounting approach. By using this approach they have discuss sources of growth or technical change
by different way. Solow approach has been discussed in previous module. The present module is relating to
Denision’s and JorgensonGriliches sources of growth. Firstly, Denison approach will be discussed.

32
 A number of studies for United States have done by Denison. He have identifies a number of sources of
growth and estimates the portion of the growth rate attributable to each. He divides the sources of growth into
four important categories;

I. The contribution of two factor of production i.e., labour and capital, adjusted for quality changes
but not depend on technical change.
II. Advancement in knowledge, which is a true measure of total factor productivity (TFP), obtained as
a residual.
III. Resource allocation improvement
IV. Economies of scale

Measurement of Sources of Growth: Empirical Explanation

In his study entitled ‘The Sources of Economic Growth in the United States’, he estimated the
contribution of different sources with the help of Cobb-Douglas type production function. He kept all inputs
i.e., labour and capital together. Denison marked an index of the stock of inputs on the basis of the base year
1929. For constructed this weighted index, he used relative share of income in the base period. In calculating
the contribution of education to output, he has treated workers of different educational categories as different
inputs. Then the growth rates of the number of workers in different educational categories were aggregated into
an index of the growth rate of total labour input according to their shares of total labour hours.

For index of capital, he was taken four types of capital inputs;

• Non-farm residential structures

• Other structures and equipment
• Inventories and;
• US international assets

Each types of capital was weighted by its own base year returns in the estimating its contribution to
growth.

The index of the contribution of increases in output per unit of input comprised advancement in
knowledge, resource shift from agriculture to industry and economies of scale

Table 1: Sources of Growth of Real National Income of the US during 1929- 57

Sources of Growth Growth Rate Per cent of Growth Rate

A. Real National Income 2.93 100.0
B. Increase in Total Inputs 2.00 68.3
(a) Labour 1.57 53.6
(b) Capital 0.43 14.7
(c) Land 0.00 0.00
C. Increase in Output per Unit Input 0.93 31.7
(a) Advance in Knowledge 0.59 20.1
(b) Resource Shift 0.07 2.4
(c) Economies of Scale 0.34 16.6
(d) Irregular Factors 0.07 -2.4
33
 The above table 1 shows that the growth rate of real national income was 2.93 per cent per annum
during the period 1929-57 in the United States economy. The growth rate is calculated from the real net
national product at the factor cost. Out of the 2.93 per cent growth rate, 2 per cent are accounted by increase in
total inputs and 0.93 per cent increase by productivity (increase in output per unit of input). Out of 2 per cent
increase in inputs, 1.57 per cent is accounted by labour and 0.43 per cent by capital.

The contribution of productivity of US real income growth was 0.93 percentages. That is 32 per
cent of total sources of growth. According to Denison, this is the residual factor. He divides the increases in
output per unit of input into three main components i.e., advance in knowledge, resource shift from agriculture
to industry and economies of scale. Thus of the 32 per cent contribution of increase in productivity is comes
from advance in knowledge (20 per cent), resource shift (2.4 per cent) and economies of scale (12 per cent).

According to Denison, the contribution of education increases the quality of labour force while
the advancement in knowledge is a technical change. Denison regards advancement in knowledge as the ‘true
residual’ and education as ‘guesstimated’. So far as other factors like recourse shift and economies of scale are
concerned as the lower size of true residual.

Criticisms of Denison’s Approach

Denison approach of sources of growth is different from Solow’s approach of sources of growth. R.
Solow attributes the residual to technical change on the other side Denison breaks the residual into further
components. Denison attributes increases in growth to improvement in the quality of labour force as a
consequence of better and more education and advancement in knowledge.

However, he has been criticized for the following weakness in his study of sources of growth;

❖ Economists have questioned the effect of education on earnings which is the index of quality of labour.
They find the adjustment factor of 40 per cent for ability, leaving 60 per cent of differences on income
differentials.
❖ Lundberg has criticized the use of Cobb-Douglas production function by Denison for calculating the
contribution of factors of production to growth rate of national income. According to him, the specific C-D
function attributes large share of labour income and low share of capital. He also points out that a static
equilibrium concept like the production function is a doubtful tool for analysis the dynamics of growth.
❖ Denison’s estimates are based on constant returns to scale, which are available after making payments to all
factors according to their marginal productivity. This is unrealistic assumption.
❖ Denison has also been criticized for assuming ‘disembodied technical’ progress. In fact, this process should
be ‘embodied’ in plant and equipment. According to J. Sandee, ‘the believer of ‘embodied’ progress
usually finds at least twice the yield deduced by the classical Cobb-Douglas disembodied trend analyst,
because he consider the whole ‘residual’ as the result of new investment’.
❖ Denison does not consider joint effects of capital and technology. Rather, he treats them as separate
elements and does not attribute technical progress to the extra capital.

Concluding Remarks

❖ Despite the above criticism, Denison has performed an extremely useful work in quantifying the
contribution of increases in physical inputs i.e., education and advance of knowledge, to growth.
34
 ❖ His estimates of sources of growth can be accepted with some level of confidence.
❖ He attempts to quantify the sources of increases in output per unit of input,
❖ However, some of his conclusions must be considered of doubtful worth
❖ The residual factor in economic growth remains the coefficient in our ignorance.

Jorgenson-Griliches Sources of Growth

Introduction

Jorgenson-Griliches examined a hypothesis concerning the explanation of changes in total factor

productivity (TFP) in his study of sources of economic growth in the United States private domestic sector
economy. According to this hypothesis, if the measurements of quantity of capital and labour are accurate then
the growth in total output is mainly accounted by growth in total inputs. The differences between the rate of
growth of real product (output) and real factor input is called the rate of growth of total factor productivity
(TFP). In the social accounting framework, it is hypnotized that if the real product and real factor input are
accurately measured then the observed growth in total factor productivity is negligible.

Jorgenson-Griliches has pointed out that there are many sources of error in system of social accounting
of real product and real factor inputs. The error is frequently creeping into the measurement of movements in
total factor productivity, which biases the estimates upwards. There are following four important sources of
errors;

❖ Errors in aggregation in combining in goods (investment and consumption) and services (labour and
capital)
❖ Errors of measurement in the prices of investment goods
❖ Errors from assuming that the flow of labour and capital services is proportional to stocks of capital and
labour.
❖ Errors as a result from the aggregation of investment goods and capital services on the side and on the
other side of labour services.

Measurement of Sources of Growth: Empirical Explanation

Jorgenson-Griliches have construct indices of total output and total input for United States domestic
private sector economy for the period 1945-65 without correcting for error of measurement to prove their
hypothesis. They have taken the US domestic private sector economy in constant prices for an initial index of
total output. They have also taken the sum of labour and capital services in constand prices for construct index
of total input. Labour and capital services are assumed to be proportional to stocks of labour and capital
respectively. The number of persons engaged in the private domestic sector of Unites States economy is taken
as stock of labour and the sum of land, plant, equipment and inventories employed in this sector is taken as
stock of capital. The difference between the rate of growth of total output and total input is called the rate of
growth of total factor productivity (TFP).Jorgenson-Griliches found that the average growth rate of total output
was 3.49 per cent per annum during the period 1945-65 and the average growth rate of total input was 1.83 per
cent per annum during the same period. Then the average growth rate of total factor productivity was 1.60 per
cent per annum. Thus, the contribution of total input and total factor productivity (TFP) in total output growth
was 52.4 per cent and 47.6 per cent respectively. These were the initial estimates of growth rate of total output,
total input and total factor productivity
35
Elimination of Errors

After these initial estimates of growth rate of total output, total input and total factor productivity,
Jorgenson-Griliches eliminate the errors of aggregation and measurement. They have reached the following
estimates shown in Table 2.

Table 2: Total Output, Total Input and Total Factor Productivity of US during 1945- 65

Average Annual Growth Rates (%) Contribution Contribution

Estimates of Input to of TFP to
Output Input TFP Output (%) Output (%)
A. Initial 3.49 1.83 1.60 52.4 47.6
After Correction For;
B. Errors of
3.39 1.84 1.49 54.3 45.7
Aggregation
C. Errors in
Investment 3.59 2.19 1.41 61.0 39.0
Goods Prices
D. Errors in
Relative 3.59 2.57 0.96 71.6 28.4
Utilization
E. Errors in
Aggregation of 3.59 2.97 0.58 82.7 17.3
Capital Services
F. Errors in
Aggregation of 3.59 3.47 0.10 96.7 3.3
Labour Services

Errors of Aggregation: After elimination of error of aggregation of consumption and investment

goods Jorgenson-Griliches found that the average growth rate of total output was 3.39 per cent per annum
during the period 1945-65 and the average growth rate of total input after elimination of labour and capital
services was 1.84 per cent per annum. Then the average growth rate of total factor productivity was 1.49 per
cent per annum. Thus, the contribution of total input and total factor productivity (TFP) in total output growth
after error elimination was 54.3 per cent and 45.7 per cent respectively.

Error in Investment Goods Prices: After elimination of errors of measurement of investment goods
prices, the role of total factor productivity has been decline. The above table reveals that with the errors of
measurement of prices of investment goods eliminated, the growth rate of total input to total output is 61 per
cent per annum, leaving 39 per cent per annum due to total factor productivity.

Errors in Relative Utilization of Labour and Capital Stock: Due to eliminating the errors in the
measurement of relative utilization of labour and capital stock, the growth rate of total output, total input and
total factor productivity are found 3.59 per cent, 2.57 per cent and 0.96 per cent per annum respectively. Thus,
the contribution of total input and total factor productivity (TFP) in total output growth was 71.6 per cent and
28.4 per cent respectively.

36
 Errors in Aggregation of Capital Services: When the errors in aggregation of capital services are
eliminated then the growth rate of total output, total input and total factor productivity are found 3.59 per cent,
2.97 per cent and 0.58 per cent per annum respectively. With these errors eliminated total input explains 82.7
per cent of the growth in total output, leaving 17.3 per cent per annum due to total factor productivity.

Errors in Aggregation of Labour Services: When the errors in aggregation of labour services are
eliminated then the growth rate of total output, total input and total factor productivity are found 3.59 per cent,
3.47 per cent and 0.10 per cent per annum respectively during the study period. It means that if we eliminated
the errors in aggregation of labour services then the role of total factor productivity will be insignificant.

Concluding Remarks

Thus after the removal of aggregation and measurement errors, Jorgenson-Griliches found that 96.7 per
cent per annum rate of growth of the United States private sector economy output over the study period is
explained by the growth in input, leaving 3.3 per cent due to change in total factor productivity or residual. The
latter is in marked contrast to 47.6 per cent before correction of data

Criticisms of Jorgenson-Griliches Approach

❖ As comparison of Denison, Jorgenson-Griliches present more realistic estimates of the sources of

growth of the United States. They have corrected all sources of errors while Denison corrects only for
errors in the measurement of labour services. Jorgenson-Griliches have shown that the residual or
change in total factor productivity is very small as comparison to Denison due to advance in
knowledge.
❖ However, some economists do not accept the Jorgenson-Griliches hypothesis, when the latter attribute
virtually the whole of measured growth to increases in factor of production i.e., inputs.
❖ Denison point out that their extremely low estimate of change in total factor productivity is almost
entirely due to the wholly unwarranted adjustment to the capital utilization series
❖ Further, Denison claims that there are very little difference between the results of Jorgenson-Griliches
and traditional estimates of the total factor productivity growth, which is accounted by the removal of
errors in the output series as claimed by Jorgenson-Griliches.
❖ Jorgenson-Griliches themselves indicate that the most serious weakness of their study is relative
utilization of capital and labour to adjust capital and labour input to year-to-year variation as a result of
discrepancies between them.

Summary

❖ The standard growth accounting exercises generate a Solow residual, which is typically viewed as a
measure of technological progress.
❖ Recent theories of endogenous growth allow for a sharper perspective on this residual. Specially, the
residual can be clearly interpreted within settings that allow for increasing returns and spillovers or in
models in which technological progress is generated by purposeful research. These interpretations
provide guidance for explaining the residual in terms of R&D outlays, public policies, and other
factors.
❖ The standard growth-accounting exercises provide useful information within the context of modern
theories of endogenous growth and that the recent theories can be used to extend the usefulness of
37
 traditional growth accounting. Hence, the older and newer approaches to economic growth are
complementary.
❖ Several economists such as Denison , Kendrick, Jorgenson and Griliches and others have tried to
quantify and break down the residual in to further several components. They contend that the residual is
not a catch-all and that changes in output are due to changes of quantities and qualities of inputs, in
economies of scale and advances in knowledge rather than the results of technical change, assuming a
stable production function.

MONEY IN ECONOMIC GROWTH AND TOBIN MODEL

Introduction

In neo-classical framework, the equilibrium marginal productivity of capital and rate of interest are
determined by technology and saving behavior and it is assumed that whatever is earned as profits is saved and
then invested. Such propositions automatically assume that the warranted growth rate is just equal to the
natural growth rate of the economy. However, if there is divergence between the two, the unattractively low
rates of return may create a problem for availability of investible resources. This can be rationalized if there are
other competing alternatives where people may wish to supply their savings if the economy is not providing
sufficient returns. James Tobin states that most of the models are non-monetary in nature and take only one
destination of the savings by the people and that is investment. However, in any economy there can be a
number of other portfolio choices in which people can invest their money. Hence different questions or issues
of equilibrium growth path arise when monetary assets are available to compete with ownership of real goods.

Economic Growth – The Non-Monetary Framework

In a non-monetary model of economic growth, savings are always equal to investment and this process
goes on and on until savings and investment augment the capital stock faster than the effective supplies of
other factors. Under the conditions of diminishing marginal productivity of capital, an increase in capital stock
leads the returns on capital to zero or below. At low or negative yields people reduce or discontinue their
savings or even to consume capital. The responsiveness of savings to rate of interest actually determines the
upper limit to capital deepening and a lower bound to the rate of return on capital. But lower savings in the
classical framework do not lead to a situation of under employment in the economy as fall in savings means
increase in consumption expenditure which automatically stimulates investment and the economy is back on
the track. This can be illustrated with the help of the figure 1. In this figure, on horizontal axis, capital intensity
i.e. k, which is defined as the quantity of capital per effective manhour of labour. Here the term ‘effective
manhour’ actually takes in to account the improvements in the quality of labour inputs due to labour
augmenting technological progress. The vertical axis measure rates of growth, rate of return etc. In the figure,
AA´ represents ‘y’ the average annual product of capital. As the units of capital increase, the average product
per unit of capital increases, therefore, this curve is downward sloping. It is also the reciprocal of capital-output
ratio. The curve MM´ represents the marginal product of capital. Since, we assume the diminishing marginal
productivity of capital, this curve is also downward sloping and becomes zero and even negative for more
intense use of capital. Both the average and marginal product curves here are shown in net terms i.e. net of
constant rate of depreciation δ. The curve S1S1´ reflects saving behavior. It shows net savings and investment
per unit of the existing capital stock. This tells about the speed of growth of the capital stock. In Harrodian
frame work it may be stated as ‘warranted rate of growth’ of capital stock. The effective labour force grows at
38
a constant rate ‘n’ independent of the capital intensity and since the ‘natural growth rate’ of the economy
depends on the natural increase in the labour force, this will be represented by a horizontal line NN´. If the rate
of growth of capital is just equal to ‘n’, there will be no change in capital intensity. If the warranted rate of
growth of capital exceeds the ‘natural’ growth of labour ‘n’, then capital deepening will occur. If capital grows
more slowly than labour, ‘k’ will decline. These facts are indicated in the diagram by the arrows in curve
S1S1´ where, the equilibrium capital intensity is k1 to which the corresponding marginal product is M1 which
is negative in this diagram. If we assume a constant propensity to save and consume, we could have remained
on the curve S1S1´ but if we consider the change in propensity to consume and save, this behavior will be
depicted by S2S2´. Here, the ratio of net investment to output declines with ‘k’. This may happen due to a
number of facts such as due to lower yield on savings, the propensity to consume increases and savings falls.
Tobin says, this may also happen as the capital deepening also implies an increase in wealth relative to current
income as a result the savings as a proportion to average income fall. In the figure, with the saving behavior

S2S2´, the equilibrium capital intensity is k2 and marginal product is M2. This is how the classical
model in non-monetary framework works.

Introducing the Monetary Factors

Some growth models assume a lower limit on the marginal product of capital e.g. Harrod argues that
investors will not simply invest unless they do not receive a certain minimum rate of return but the savers are
not discouraged from trying to save when yields fall to or below this minimum. In Keynesian framework such
situation may lead to deficient demand and unemployment while Harrod says that the difficulties arise when

39
the warranted rate of growth at the minimum required rate of profit exceeds the natural rate. The rate of
savings from full employment output would cause capital to accumulate faster than the labour force is
growing. As a result, the marginal product of the capital would fall and push the rate of return on investment
below the required minimum. In figure HH´ is the required minimum. Then the kH is the maximum capital
intensity investors will tolerate. Yet both the saving behaviors S1S1´ and S2S2´ show that the economy will be
pushed towards marginal products M1 or M2 and the capital intensities k1 or k2. It is this excess of ex ante
savings over investment which gives rise to the Keynesian difficulties. The opposite problem would arise if the
warranted rate of growth falls short of the natural growth rate, in that case higher return on investment would
raise the demand for investment goods well above savings. This impasse will lead to ‘inflationary gap’ in the
economy whose main consequence is an increase in general price level. The increase in prices is another form
of forced savings which reallocates the income from low savers to the high savers and ultimately, the
discrepancy in savings and investments is removed. But it is nearly impossible to talk about inflation in a non-
monetary model. Moreover, a minimal rate of return on capital can not exist in isolation, it must reflect the
competition of other channels for the placement of savings. Therefore, we may consider other alternatives of
store of value i.e. monetary assets. It is their yield that set limits on the acceptable rates of return on real capital
and on acceptable degree of capital intensity. Therefore, it is essential to introduce the monetary assets. Tobin
takes the case of a single monetary asset with the following properties:

1. It is supplied by the central government. This means that it represents neither a commodity produced by
the economy nor the debts of private individuals or institutions.
2. It is the mean of payment, the medium of exchange and a store of value by reason of its general
acceptability in the discharge of public and private transactions.
3. Its own yield is fixed by the government arbitrarily. This may even be zero.

Assumptions

The introduction of monetary assets in the growth model needs certain assumptions. These are
discussed below:

1. The value of money in terms of goods is fixed.

2. All the individuals make rational choices.
3. There is perfect information about the returns on available alternatives of store of value.
4. The community’s wealth has two components – the real goods and paper goods.
5. The monetary assets can be easily converted in to physical assets and vice versa.

The Explanation of the Model

For a rational individual, in case of two alternative portfolios, the wealth owners will wish to place all
of their wealth in the assets with the higher yield. If they are the same, the wealth owners do not care in what
proportion they divide their wealth between the two assets. If there are positive supplies of the both assets, they
will hold their portfolios in such a way that the two yields are equal. This is only through this behavior that we
can understand how the institutionally determined rate of interest on money controls the yield of capital. In
particular, it is this rate of interest which is the minimal rate of profit that leads to the deflationary impasse.
Now by introducing the central government assets, the government can avoid this impasse in two ways.
Returning to figure 1, if we take HH´ as the yield on money, it will be the minimal yield acceptable to the

40
owners of capital. The corresponding capital intensity is kH. One measure the government could take it to
reduce the yield on money, say to M1. Alternatively, the government could channel part of the community’s
excessive thrift in to increased holdings of money. The Harrod impasse occurs if the saving behavior is S1S1´
but if only part of it goes to capital accumulation, then the rate of increase in capital stock can be lowered to
S3S3´ then there will be no problem as at this curve, the equilibrium capital intensity will be kH, consistent
with maintaining the marginal product of capital at the required level HH´. This can be done if the government
provides new money to absorb the saving represented by the difference between S1S1´ and S3S3´. The only
way for the government to achieve this is through continuously running deficit finance by issue of new money.
But we also have to find the optimal size of this deficit. This can be observed from figure 2.

In figure 2 the vertical axis shows the savings and the horizontal axis shows the income. Both of these
variables are shown in terms of per unit of capital. yH is the output per unit of capital corresponding to the
required equilibrium in capital intensity kH. The government purchases of goods and services are assumed to
be a fraction ‘g’ of output. Consequently, yH(1 – g) is output available for private use and is also the
disposable income of the population. Taking S1S1´ as the saving function, SH is the amount of private saving
(relative to capital stock) when the budget is balanced. However, this much of the investment causes the
warranted rate of growth to exceed the natural rate. Now ‘n’ is the natural rate of growth, it is therefore, the
right amount of investment. A deficit of dH will probably do the trick. It will increase the disposable income
from yH(1- g) to yH(1-g) +dH, and this raises total savings to SH´. But out of this dH is the acquisition of
government debt, leaving only ‘n’ for new tangible investment. This can also be expressed mathematically, the
total savings would depend upon the propensity to save i.e. ‘s’ out of this increased disposable income and can
be written as:

S = s[y(1-g)+d] = d+n ------(1)

It can also be written as

41
𝒅 𝒔(𝟏−𝒈)−𝒏/𝒚
= -------- (2)
𝒚 𝟏−𝒔

Since ‘d’ stands for deficit, the equation (2) gives the required deficit as a fraction of income. This shows that
the absolute size of the government debt is immaterial, it can be of any size and it should be a given proportion
corresponding to the given level of income.

Similarly, we can also discuss the case of inflationary impasse, in which contrary to the above case, the
government will adopt the surplus policy. In this case, a balanced budget policy would leave the yield on
capital so high that no one wants to hold money. To get public to hold money, it is necessary to increase the
capital intensity and lower the marginal product of capital. But a higher capital intensity takes more investment
relative to output. To achieve a higher investment ratio, the resources that savers make available for capital
formation must be supplemented by a government budget surplus. This process can be seen in the figure 2
itself but by reading it in reverse manner

Above discussion shows that the economy will remain at the equilibrium level. Under such an
equilibrium, the shares of money and capital in total wealth must be constant so that their yields can remain
constant. To maintain the fixed relations between the stocks, money and capital must grow at the same rate i.e.
new savings must be divided in the same ratio of the portfolios as the old savings. For further elaboration, let
m(k, r) be the required amount of money per unit of capital when the capital intensity is ‘k’ and the yield of the
money is ‘r’. since, ‘m’ is an increasing function of ‘r’; more money will be demanded when its yield is higher.
If we take ‘r’ as fixed, ‘m’ will be increasing function of ‘k’ because an increase in’k’ lowers the yield of
capital, it also lowers ‘y’ and therefore the transaction demand. Let ‘w’ be the warranted rate of growth of the
capital stock, and let ‘d’ is deficit per unit of existing capital as already defined in equation (1) and (2). The
constancy of amount of money per unit of capital at m(k.r) requires that d = m(k,r)w. Assuming as before that
savings are constant proportion of disposable income, the basic identity in equation (1) can be rewritten as

S = s[y(1-g)+d] = d+w

And taking d = m(k,r)w; the equation (2) becomes

𝒔𝒚(𝒌)(𝟏−𝒈)
𝒘(𝒌, 𝒓) = --------- (3)
𝟏+(𝟏−𝒔)𝒎(𝒌,𝒓)

In equilibrium, w = n, the equilibrium degree of capital intensity is the value of ‘k’ that equates w(k,r) in above
equation with ‘n’. In this equation ‘w’, ‘m’ and ‘y’ are written as functions of ‘k’. Since, ‘y’ is decreasing and
‘m’ is an increasing function of ‘k’, it is but natural that ‘w’ declines with ‘k’. Moreover, the amount by which
‘w’ in (3) falls short of the hypothetical ‘w’ for m=0; s.y (1-g) increases with ‘k’. This may be analysed in
figure 3.

42
 Like in figure 1, in this figure also, we have S1S1´ as a saving function where savings are a constant
fraction of disposable income. This would be warranted rate of growth if m=0. If we assume that the stock of
money is adjusted to the capital intensity by deficit financing, the W1W2´ represents the warranted rate of
growth of capital for every capital intensity. The intersection of W1W2´ with natural growth rate curve i.e.
NN´, the equilibrium capital intensity will be k1 with marginal product M´. This yield, however, is not
necessarily equal to the yield on money ‘r’. The curve W1W2´ is drawn for a particular yield on money r̅1
lowering the yield on money, say to r̅2 would shift the curve to the right to W2W1´ increasing equilibrium
capital intensity and lowering the equilibrium rate of return on capital.

Going a step further, let us now suppose that the value of money in terms of goods is variable. This will
have two important consequences – the real value of the monetary component of wealth is not under the direct
control of the government but also depend upon the price level; and the real return on a unit of money will
consist not only its own yield but also the change in its real value. Thus, under the conditions of deflation, even
if the nominal stock of money remains constant, the increase in real value of the money stock will play the
same role as in case of deficit financing as it will increase the real disposable income and it will absorb part of
the propensity to save. However, in the equilibrium, the real stock of money must be increasing as fast as the
capital stock, namely at the natural rate ‘n’. in the present case, this can happen only if the price level falls at
rate ‘n’ and the real return on money ‘r’ is not simply the nominal yield r̅ but r̅+ n. Consequently, the demand
for money will be larger than if the prices were expected to remain stable. Equilibrium will require a greater
stock of money per unit of capital and a lower capital intensity if deflation is substituted for money creation.
This can be observed from figure 3 where W3W3´ is the curve corresponding to a yield on money ‘n’ points
higher than the yield in case of W1W2´.

43
Equilibrium Path and its Stability

Now it is important to discuss the stability of the equilibrium growth path when the community is
thrown out of portfolio balance due to irregularity in technological progress, labour force growth, saving
behavior, change in yield expectations, or portfolio preferences. If people have too much capital and too little
money for its tastes, goods etc. the prices will fall faster or rise more slowly than before. In the opposite case,
the public will try to buy capital with money and will push the prices up faster or fall slowly. Under such
conditions, the issue of stability has been addressed in figure 4.

In this figure, the vertical axis measures the price deflation i.e.− 𝑝 ̂ 𝑃 and the horizontal axis measures
the rate of capital accumulation i.e. K̂ /K. On each axis the natural rate of growth i.e. ‘n’ has been shown. On
the horizontal axis, a rate of accumulation larger than ‘n’ means capital deepening and a decline in yield, while
opposite will be the case if capital accumulation is at a slower rate than ‘n’. The 45° line from the origin,
labeled ‘portfolio balance’ shows the combinations of price deflation and capital formation that will preserve
portfolio balances at the existing rates of return. The negatively sloped line labeled savings shows the
combinations of –p̂/p and K̂ /K that exhausts the savings, assuming the same saving behavior as assumed in
figure 2. On the horizontal axis, s.y(1- g) is the rate of capital growth if all savings go into capital. On the
𝑆 𝑝.𝑦
vertical axis1−𝑆 . measures a rate of price deflation at which the entire propensity to save would be satisfied
𝐷
by capital gains on monetary assets, here D is the nominal stock of money per unit of capital. The saving line
crosses the portfolio balance line at the point (n, n). At this point, new savings will be divided so as to maintain
both portfolio balance and capital intensity. But if deflation exceeds ‘n’, the division of saving will move to the
north-west along the saving line. This means that the yield on money will be higher but the portfolio behavior
will not adjust it immediately and will take time until the new rate of deflation is registered in the expectations
of the wealth owners to make adjustments according to new expectations. Meanwhile, the yield on capital will
be rising because capital accumulation is falling short of the natural rate. Further, as the ratio of income to
44
money stock (py/D) is declines, the vertical intercept of the saving line moves down; the rate of deflation that
would divert all savings away from capital formation becomes smaller and smaller. So the yield on money
declines, the yield on capital rises, while the relative supplies will move in the opposite direction. Eventually,
the rate of deflation will fall to ‘n’ and the economy will be on balanced growth path. However, a great
hindrance to achieve this path can be the downward rigidity of the money wages which may prevent the
reallocation of the portfolios that move towards the balanced growth.

Summary

In the classical theory the interest rates and the the capital intensity of the economy are determined by
the interaction of the technology and the saving propensities both in the short as well as the long run. Keynes
gave reasons why in the short run monetary factors and portfolio decisions modify, and in some circumstances
dominate, the determination of the interest rate and process of capital accumulation which influence the growth
path of the economy. James Tobin suggested that similar proposition is true even in the long run. The
equilibrium interest rate and the process of capital accumulation may in general be affected by monetary
supplies and portfolio behavior as well as by technology and savings. Tobin states that if under the conditions
of low returns, people are averse to invest, there can be other alternatives that have the capacity to plug this
gap. He proposed monetary debt of the government as one alternative store of value and show how enough
savings may be channeled in to this form to bring the warranted rate of growth of capital down to natural rate.
The equilibrium capital intensity and interest rates are then determined by portfolio behavior and monetary
factors as well as saving behavior and technology. In such an equilibrium, the real monetary debt grows at the
natural rate, either by deficit spending or by deflation leading the economy to a stable equilibrium.

LEVHARI, PATINKIN AND JOHNSON MODELS

Introduction

Recently, most of the economic theorists have realized that money has an important role to play in the
real growth models. For this purpose, the monetary growth models were introduced which shed some light on
matters of monetary policy in modern growing economies. The monetary growth theories have large
overlapping dual role – first, to construct a theoretical framework which would allow for a meaningful
integration of monetary theory and real growth theory, and second, to provide meaningful suggestions on
question of monetary policy. D. Levhari and D. Patinkin (1968) and H.G. Johnson (1962) objected to Tobin's
model on the basis of the fact that it treats money solely as a store of value and, in effect, ignores the services it
performs in overcoming transactions costs, etc. In the works of Levhari and Patinkin's, it has been argued that
the existence of cash balances in an economy facilitates production. Levhari and Patinkin hold the view that
money is held only because it enables the economic unit in question to acquire or produce a larger quantity of
commodities in the usual sense of the term. They discussed money as a producer good and as a consumer good
and it can also play both these roles simultaneously. On the other hand, Johnson has indicated the strengths and
weaknesses of the classical, Keynesian and post-Keynesian versions of money demand and their impacts on
output and income.

Levhari and Patinkin Model

David Levhari and Don Patinkin (1968) introduced a model of economic growth in which money is
treated as consumer’s good. They demonstrated that if the aggregate price level changes at a constant rate and
45
the demand for real balances with respect to the money rate of interest is inelastic, then the balanced growth
path is also stable.

The model consists of the following equations:

Production is function of labour and capital:

Y = F(K, L) = Lf(k) (1)

lim lim ′ (k)

f ′ (k) > 0, 𝑓 ′′(k) < 0, f ′(k) = ∞; f =0
k→0 k→∞
It is homogenous of degree 1 in capital (K) and labour (L) and has the usual neo-classical properties.
The per capita output i.e. Y/L can therefore be expressed as a function of capitallabour ratio ‘k’ and labour
supply grows at a constant rate ‘n’.

𝑘 ≡ 𝐾/𝐿 (2)

Saving ‘S’ is a constant fraction of disposable income. The assumption of constancy of the saving rate
can be related with nominal money supply which grows at the rate 𝜇, and ‘P’ is the price level which changes
at the rate 𝜋. The real rate of return on capital (r) is the marginal product of capital i.e. 𝑓 ′ (𝑘). If we do not take
in to account the utility yield of holding money balances, then the saving function is:
𝑀
𝑆 = 𝑠 [𝑌 + (𝜇 + 𝑟) (3)
𝑃

Disposable income is obtained by as the output ‘Y’ plus the real value of government transfer payments
𝜇𝑀 𝜋𝑀
less the loss in real value of existing cash balances due to inflation plus the opportunity cost of holding
𝑃 𝑃
𝑀
money which is 𝑟 + 𝜋 ( 𝑃 ). Not all of the savings go in to investment. The rate of change of real balances
𝑀
i.e. 𝑃 (𝜇 − 𝜋) can not be held in the form of physical assets. Hence, under the conditions of equilibrium in
capital market, ‘I’ the investment function and can be defined as:
𝑀
𝐼=𝑆− (𝜇 − 𝜋) (4)
𝑃

̂ = 𝐼 − 𝛿𝐾
𝐷𝐾 = 𝐾 (5)

𝑟 = 𝑓 ′ (𝑘) (6)
𝐷𝐿 𝐿̂
=𝐿=𝑛 (7)
𝐿

𝑀/𝑃𝐾 ≡ 𝑚 = 𝜆(𝑖)𝑓(𝑘)/𝑘 (8)

𝑖 =𝑟+𝜋 (9)
𝐷𝑀 ̂
𝑀
=𝑀=𝜇 (10)
𝑀

𝐷𝑃 𝑃̂ ̂
𝑀
=𝑃 = 𝜋 =𝜇−𝑛−𝑀 (11)
𝑃

46
 Total consumption is now represented by (1 – s)[F(K, L) + (M/P)(µ+r). But according to Levhari and
Patinkin, in order to obtain equality of savings and investment, we need to deduct from total output, not the
total consumption expenditure but the physical consumption. Hence, the change in capital stock is,

̂ = 𝐹(𝐾, 𝐿) − {(1 − 𝑠) [𝐹(𝐾, 𝐿) + 𝑀 (𝜇 + 𝑟)] − 𝑀 (𝑟 + 𝜋)}

𝐾 (12)
𝑃 𝑃

It is further assumed that the price level instantaneously adjusts itself so as to equate the demand and
supply for real balances, that is

𝑀𝑑 𝑀
= 𝜆𝐹(𝐾, 𝐿) = (13)
𝑝 𝑃

The desired level of physical consumption, Cp, should be represented not by the second term of
equation (12) but by:

𝑀 𝑀𝑑
𝐶𝑝 = (1 − 𝑠) [𝐹(𝐾, 𝐿) + (𝜇 + 𝑟)] − (𝑟 + 𝜋) (14)
𝑃 𝑃

Thus, the disposable income which determines the desired total consumption, the relevant quantities are
𝑀
the volume of the transfer payments actually received by individuals i.e 𝜇 𝑃 imputed income from real balances
𝑀
actually held (i.e. 𝑟 𝑃 Individuals always try to achieve the desired balance in their real balances which they
consider to be convertible in to physical capital at ‘r’ rate of return. On the other hand, the desired level of
consumption, is determined by the amount of liquidity services which individuals have. The real balance effect
of equation (14) can be derived by differentiating it partially w.r.t. (M/P).

𝜕𝐶𝑝
=𝜇+𝑟
𝜕(𝑀/𝑃)

Levihari and Patinkin have argued that any rationale for holding money is actually to use it to purchase
the consumer’s goods or producer’s goods. If money is treated as a consumer’s goods then it means that people
derive utility from money holdings as money provides protection against uncertainties. However, there is some
opportunity cost of holding money which is expressed in terms of nominal rate of interest (i) which is equal to
real rate of return (r) plus the rate of inflation (𝜋).

The demand for real balances can be written as:

𝑀𝑑
= ℎ(𝑖. 𝑦), ℎ𝑖 < 0; ℎ𝑦 > 0
𝑃𝐿
This indicates that the demand for real balances is negatively associated with nominal rate of interest
and positively associated with the level of output. Other things remaining the same, if the rate of return on
capital and/or rate of inflation increases, the demand for real balances falls. Since money supply is growing at
the rate ‘𝜇’, total money supply in the economy consists of
𝑠
𝑀(𝑡) = 𝑀𝑜 exp (𝜇𝑡)

47
 ̂
𝑚
Thus, the expected cost of holding money balances is equal to 𝑟 + 𝜇 − 𝑛 − 𝑚. Here, the rate of inflation
̂
𝑚
is equal to 𝜇 − 𝑛 − 𝑚

𝑀𝑑 𝑀𝑠
The symbols 𝑚d and 𝑚s are equal to 𝑎𝑛𝑑 respectively. It is assumed that money market is in
𝑃𝐿 𝑃𝐿′
equilibrium and 𝑚d = 𝑚s . Further, it is also assumed that it is the price level that brings this equality. The
excess demand in the commodity market is equal to the gap between planned investment and planned savings
in goods.

𝐼 𝑆
( − ) + 𝜔(𝑚𝑑 − 𝑚𝑠 ) = 0
𝐿 𝐿
Here, ‘𝜔’ is the adjustment factor and assume that the excess demand for asset flows is a constant
proportion to excess demand for asset stocks and the planned investments are assumed to be equal to the
planned savings. But the change in capital stock is the difference between availability of capital and rate of
𝑀
change of real balances i.e. 𝑃 (𝜇 − 𝜋).

𝑀
̂ =𝑆−[
𝐾 (𝜇 − 𝜋)]
𝑃
̂
𝐾
and ̂ + 𝑛𝑘
=𝐾
𝐿

Using all these functional relationships, the investment function in equation (4) becomes:

̂
𝐾
= ∅(𝑘) − 𝑛
𝐾
this can also be termed as the dynamic path of capital intensity
𝑠𝑓(𝑘) 𝑚 𝑚
∅(𝑘) = − (1 − 𝑠) 𝑛 𝑘 + 𝑠𝑖 (15)
𝑘 𝑘

∅ ′ (𝑘) < 0. For steady-state equilibrium:

𝑘𝑓 ′ (𝑘)−𝑓(𝑘) 𝑑(𝑚/𝑘) 𝑑𝑖
∅′ (𝑘) = 𝑆 − [(1 − 𝑠)𝑛 − 𝑠𝑗] + (𝑠𝑚/𝑘) 𝑑𝑘 (16)
𝑘2 𝑑𝑘

Increase in capital stock has an uncertain effect on excess demand for capital but may either increase or
decrease savings since the income and wealth effects on savings work in opposite directions. If the system is
near the golden rule then the second term is positive but the first and the last terms are unambiguously
negative. Assuming the steady state path, the effective labour supply and the nominal quantity of money grow,
respectively at the constant rates that are ‘n’ and ‘µ’ where, the former is always positive, but the latter can be
negative. From constancy of capital labour ratio, it follows that:
𝑘́ ̂
𝑘 𝐿̂ ̂
𝐾
=𝑘−𝐿 =𝐾−𝑛 = 0 (17)
𝑘

48
 𝑀
In this system, the real value of per capita physical capital denoted as 𝑚 = 𝑃𝐿 is also assumed to be
constant i.e.
̂
𝑚 ̂
𝑀 𝑝̂ 𝐿̂
=𝑀−𝑝−𝐿 =𝜇−𝜋−𝑛 =0 (18)
𝑚

Thus, total physical capital and total real money balances both expand at ‘n’. Now it is important to
find the steady state value of ‘k’. This can be found by substituting (13) in to (12) and dividing both sides by
‘K’, we obtain,
𝐹(𝐾,𝐿)
{𝜆𝑛 − 𝑠[1 + 𝜆(𝑛 + 𝜋 + 𝑟)]} + 𝑛 = 0 (19)
𝐾

Using equations 17 and 18 and dividing and multiplying the first term of equation (19) by ‘L’, we get

𝑓(𝑘){𝑠[1 + 𝜆(𝑛 + 𝜋 + 𝑟)] − 𝜆𝑛} = 𝑛𝑘 (20)

The distinguishing feature of a market economy is that not all the savings need to be devoted to
augmenting the capital stock, some of it can also be used for maintaining the real balances. The savings
devoted to physical capital are:
𝑀
𝑑( )
𝑃
𝑆𝑝 = 𝑆 − (21)
𝑑𝑡

But this function of physical savings can also be written as 𝑆𝑝 = 𝑌 − 𝐶𝑝, hence, the equation (21) can
be rewritten as:
𝑀 𝑀
𝑆𝑝 = 𝑠 [𝑌 + (𝜇 + 𝑟)] − (𝜇 − 𝑟) (22)
𝑃 𝑃

In the steady state it becomes:

𝑆𝑝 = 𝑠𝑌[1 + 𝜆(𝑛 + 𝜋 + 𝑟)] − 𝜆𝑛

Both Levhari and Patinkin consider the services emanating from the holding of real money balances as
influencing individual’s disposable income. They show that when money is introduced as a consumer good in
a single sector neo-classical growth model, the effect of inflation upon the degree of capital intensity in steady
state equilibrium is ambiguous. They argue that it is the dropping of the assumption of ‘a constant saving ratio’
rather than the different definition of disposable income which yields qualitatively different results from those
of the Tobin Model.

Money can also be considered as a producer’s good i.e. it is held as it enables the economic unit in
question to acquire or produce a larger quantity of commodities. Thus, money is being considered as an
inventory. Here, the real value of money can be introduced in to the production function.

𝑌 = 𝐺(𝐾, 𝐿, 𝑀/𝑃)

This is linearly homogeneous in all variables. The relationship in this equation shows that the
production is function of labour (L), physical capital (K) as well as the real value of the working capital or real

49
money balances. But it is also assumed that money is an exogenous variable which is without any cost of
production and administration.

Actually, this theory by Levhari and Patinkin has been presented as dealing with the effects of the rate
of monetary expansion in the growth path of the system. However, the effects of a monetary expansion differ
in accordance with the use of this additionally created money. The government may use it for transfer
payments, consumption (expenditures from the current budget on goods and services) and investment
(corresponding expenditure from the development budget). Even if the government does not itself carry out
investment, it can affect the level of capital intensity in the economy by varying the proportion of its budget
devoted to the consumption of goods and services as against transfer payments. It actually increases the
physical consumption in the economy and reduces the savings and ratio of capital by transferring its
expenditure from transfer payments to consumption of goods and services by government. Actually, in this
case the money transfers from the household sector which has a propensity to consume less than one to the
government which has a unitary marginal propensity to consume. Moreover, it has been assumed that money is
issued by the government and is therefore, an outside variety. But if we consider money as an inside generation
of money through deposit creating activities of the banking sector or bonds/loans market, it is held by this
theory that the behaviour of the individuals would not be affected by either the rate of change in prices or by
the rate of expansion of the money supply. Hence, the system is neutral, especially in the second sense which
states that if nominal quantity of money is twice as high as before then a price level will also be twice high
which will enable the system to return to equilibrium with all the real variables having the same values as
before. To sum up the theory by Levhari and Patinkin, we can say that money can be used as a consumption
good as well as the producer’s good and it can play both these roles simultaneously. Hence, a general model of
demand for money should analyze it both as a utility function and production function.

Johnson’s Views on Money Demand

H. G. Johnson (1962) has given a complete analysis of Classical, Keynesian and the ISLM approach of
determination of rate of interest. He proposed that the full Keynesian theory, in which the interest rate, income,
saving and investment, demand for money and supply of money are all mutually interdependent, can be
represented by the IS-LM curves. Johnson states that although, the IS-LM approach was originally developed
by J. R. Hicks which proved very useful standard tool for monetary theory but it can also be adapted to take
account of more general assumptions that Keynes made – e.g. saving depends partly on income and that both
demands for money demand on both income and the interest rates and investment, partly on income, and that
both demand for money depend upon income and interest rate, it can be used to explain a number of important
relationships to solve several monetary problems. Johnson suggested that this can be done by combining the
Keynesian and Hicksian analysis. Let us first focus on figure 1. The figure shows the supply curve of output
(Xs) which is based on the assumption that the diminishing returns operate on variable factor i.e. labour. This
depicts the upward sloping nature of the supply curve. The curves Yts show different combinations of real
output and prices giving a downward sloping curve. The equilibrium price level and output level are given by
intersection of the supply curve with the real output curve.

50
Figure 1

In figure 2, we can see real wage earned by labour (the marginal productivity of labour) and the relation
between the real wages earned by labour and the quantity of labour employed which is equal to its marginal
product and hence shown by the MPL curve or the demand for labour curve. On the supply side, we can see
the relation between the real wage earned and the labour supplied i.e. the LS curve. Although, the classicals
say that the level of employment is determined by the intersection of the demand labour curve and the supply
of labour curve, yet in the Keynesian theory, the levels of employment and real wages are determined by
aggregate demand and the levels of output due to the technological relationship between them. If the real wage
rate is OW0, the difference in actual employment and the labour supplies i.e. L0LE shows the involuntary
unemployment. This indicates the additional output, it could produce and hence reduce the prices.

Johnson says that the chief argument of Keynes’ theory relates to the question if underemployment
equilibrium depends on the assumption of rigid wages. He says that the answer lies on effect of increase in
money supply on wages and employment. Actually, the effect of a wage cut in the system is the same as the

51
effect of an increase in money supply as with fall in wages and prices, less money is required for transactions,
more is available for speculative balances. With large quantities of money, there will be fall in rate of interest
leading to increase in investment and then employment. But the achievement of full employment depends upon
the fact if the increase in money supply is adequate enough to push the economy towards full employment.
There can be two cases in which increased money supply will not lead to full employment. These are:

1. The economy is already in ‘Liquidity Trap’.

2. The economy is so interest elastic that at positive interest rates, investment does not increase, or
savings fall, to a level consistent with full employment.

Thus, full employment will not be achieved by monetary expansion or say, a wage cut, if saving from a
full employment income would exceed investment at the lowest rate of interest, the money market will allow.

Johnson further highlights certain controversies of viewing the Keynesian theory of employment as a
theory of interest. These are:

1. Keynes himself initiated the debate if rate of interest is a real or monetary phenomenon. Johnson
says that since Keynes’ speculative demand for money is based on the relation between actual and
expected rates of interest, the real forces ought to enter into liquidity preference in shape of
profitability of investment.
2. Since in Keynesian theory, interest is a relation between the present and the future, expectations
must influence the rate as determined in the market.
3. Another debate is related with the issue if rate of interest is determined by demand and supply of
money or by the demand and supply of securities. To deal with this issue, we can divide the
economy in three markets – the markets for output, cash and securities. The sum of excess demand
in these three markets, Xg, Xm, and Xs, respectively, must be identically equal to zero.

𝑋𝑔 + 𝑋𝑚 + 𝑋𝑠 = 0

Under full equilibrium in the economy, the rate of interest equates both the demand and supply of
money and demand and supply of securities but if there is disequilibrium in the goods market, we can not say
that the rate of interest is determined in either of the two markets, unless it is assumed that the remaining
markets behave in such a way that the excess demand for output i.e. excess of investment over savings, is
financed by bonds sales and loanable funds so that when total demand for money is the economy is lumped
together, it is just equal to the total supply of the same.

On the other hand, in case of post-Keynesian versions, Johnson says that in case of inducement to
investment, the important development is the introduction of the relation between capital stock and output in
the form of accelerator or simply the capital-output ratio as a determinant of investment decision. This permits
the conversion of the static equilibrium of Keynesian system into cycles and growth models in which the fixed
capital investment can be considered as important determinant of income. The liquidity preference theory had
been improved in a number of ways particularly by extending the motives of holding money to asset holding as
well. Johnson says that in monetary theory, the main contribution of Keynesian theory has been to emphasise
the function of money as an asset, alternative to other assets, and to break the quantity theory assumption that
there is a direct connection between money quantity and aggregate demand. But this theory needs to be
adapted to suit the non-depression conditions.
52
Summary

Recognizing the importance of money in production, we can see that the approach of treating
real money balances as an input in the production function highlights the central role played by money in
production. But this is an inadequate approach at least for two reasons. Firstly, money should not be considered
as an input of the production function in the same way as the physical capital and labour. Money capital is a
catalyst. It has no direct marginal product but operates only by influencing the way in which other factors are
used. Including money in the production function precludes the analysis of how money affects the efficiency
and organization of production. The theory by Levhari and Patinkin has dealt with the effects of the rate of
monetary expansion in the growth path of the system. However, the effects of a monetary expansion differ in
accordance with the use of this additionally created money. To sum up, the theory by Levhari and Patinkin, we
can say that money can be used as a consumption good as well as the producer’s good and it can play both
these roles simultaneously and hence, a general model of demand for money should analyze it both as a utility
function and production function. On the other hand, Johnson tried to show that the Keynesian theory of
employment and output can easily be fitted in to the framework showing simultaneous equilibrium in money as
well as the real market. He said that the monetarists’ success was transitory because its central problem,
inflation was inherently less socially important than unemployment. Johnson tried to incorporate the valid
ideas of previous orthodoxy in to the new framework. He argued that the success of a new idea in creating a
revolution would depend upon the opportunity to escape from the intellectual authority of the past work by
deploying novel analytic and empirical techniques.

ENDOGENOUS GROWTH; INTELLECTUAL CAPITAL: ROLE OF LEARNING, EDUCATION

AND RESEARCH

Introduction

In majority of the neo-classical growth models, technical change is considered as an exogenous factor,
though it is an important driver to growth. However, Kaldor and Arrow proposed the idea of technology
having some endogenous component as well. Further, the Solow Model which is largely based on the idea of
residual effect of technical change and that in the long run, the economies move on the stagnant growth path
was challenged on several grounds e.g. its unrealistic result for exogenous technical progress accounting for
more than 70 per cent of output growth; prediction of convergence among poor and rich economies in the long
run; failure to explain widened growth disparities and long run imbalances in factor productivities etc. These
shortcomings lead to the birth of endogenous growth theory, which has been formally developed since mid-
1980s following the pivotal works by Romer (1986, 1990) and Lucas (1988) and which has soon become the
dominant framework for studying economic growth and development. One of the main reasons why
economists have grown interested in endogenous growth is because of an empirical puzzle. The neo-classical
model predicts that countries with low per-capita incomes grow faster than those with high income, so that
over time per-capita incomes converge. But a broader set of data indicated that the rich countries may follow
an ever increasing growth path while the poor economies move on at much slower pace leading to divergence
among these economies. It is also found that even the same type of technology and rate of investment may not
lead to similar results across the economies. This has also been found that even the same level of foreign direct
investment may lead to different results in different economies with different human and physical capital.

53
 Thus, it is important to find some endogenous factors that determine the rate of growth of the
economies.

Endogenous Growth Models: An Overview

The endogenous growth models in contrast to the exogenous growth models believe that the growth of
the economy is more influenced by the internal processes and policies, structure etc. than merely the external
factors. Endogenous growth is long-run economic growth at a rate determined by forces that are internal to the
economic system, that govern the opportunities and incentives to create technological knowledge. In the long
run the rate of economic growth, as measured by the growth rate of output per person, depends on the growth
rate of total factor productivity (TFP), which is determined in turn by the rate of technological progress. The
neoclassical growth theory of Solow (1956) and Swan (1956) assumed the rate of technological progress to be
determined by a scientific process that is separate from, and independent of, economic forces. Neoclassical
theory thus implies that economists can take the long-run growth rate as given exogenously from outside the
economic system. Endogenous growth theory challenges this neoclassical view by proposing channels through
which the rate of technological progress, and hence the long-run rate of economic growth, can be influenced by
economic factors. It starts from the observation that technological progress takes place through innovations, in
the form of new products, processes and markets, many of which are the result of economic activities. The
firms may learn from experience how to produce more efficiently, a higher pace of economic activity can raise
the pace of process innovation by giving firms more production experience. The spread of this knowledge and
consequently, the macroeconomic effects of any technical progress depend upon the spillover effects that again
depend upon the level of human capital in the society. Here lies the important difference between the
endogenous growth theories and the neo-classical growth theories which take technical progress as exogenous
factors. In this context, we can discuss here, the basic models of endogenous growth given by Romer and
Lucas.

Romer’s Model

This model challenges the basic assumption of neo-classical model of growth which refers to
applicability of diminishing returns in the long run. This model states that many economies have been able to
exhibit increasing returns to scale even in the long run. This long run increase in per capita income has been
possible not due to the external factors but due to the endogenous technical changes. Any change in technology
leads to positive effects on productivity of the factors whose combined effect may lead to increasing returns.
Romer has discussed about the spillover effects of technical change. He states that a technical change, any
innovation, new method of doing things do not remain with one individual only. He stresses that the physical
part of any new technology may be privately owned but the knowledge part soon becomes a public good.
Since, production is a social phenomenon, any new method used in a production unit soon spreads to another
units as under perfect markets, the factors are perfectly mobile and they carry their knowledge along with
them. A highly productive factor in a production unit not only himself/herself higher level of productivity but
also raises the level of productivity of his fellow beings in the production unit. Thus any investment in new
knowledge, research and development or human capital may have greater social returns as compared to its
private returns. As a result the national output would be increasing at an increasing rate and new unit of human
capital investment will yield increasing rate of return. In the simple framework of Romer’s model, we take a
single sector Cobb-Douglas type production function which has inherent assumptions of homogenous sectors
and applicability of constant returns to scale. But in order to include the spillover effects of technology on
54
aggregate production of the economy, it includes a separate variable of human capital. Therefore, the
production function in Romer’s model can be read as follows:

𝑌 = 𝐴. 𝐾 𝛼 . 𝐿1−𝛼 . 𝐾𝛽

Here, ‘A’ is the efficiency parameter of given level of technology, ‘L’ stands for units of labour, ‘K’ for
units of capital which is presented here both as the physical as well as

human capital. ‘Y’ is the level of output and α and β are the output elasticies of the respective variables.
Above equation cn also be written as

𝑌 = 𝐴. 𝐾 𝛼+𝛽 . 𝐿1−𝛼 --------------- (1)

At any point of time, the change in output would be possible only due to change in physical as well as
human capital and units of labour along with the changes in factor productivity, therefore,
dy ∂Y ∂K ∂Y ∂L
= . + . ------------------ (2)
dt ∂K ∂t ∂L ∂t

The value of different components of this equation can be found by differentiating equation (1) partially
𝑑𝑦 𝜕𝐾 𝜕𝐿
with respect to ‘K’ and ‘L’ and if we further denote 𝑎𝑠 ∆𝑌, 𝜕𝑡 𝑎𝑠, ∆𝐾 𝑎𝑛𝑑 𝜕𝑡 𝑎𝑠 ∆𝐿, equation (2) takes the
𝑑𝑡
following form:
∆𝑌 ∆𝐾 ∆𝐿
= (𝛼 + 𝛽) + (1 − 𝛼) ----------------- (3)
𝑌 𝐾 𝐿

∆𝑌 ∆𝐾
If both output and capital stock grow at the same and constant rate, then = = 𝑔 and since the
𝑌 𝐾
growth of labour force depends upon the growth of population which grows at a natural rate determined
∆𝐿
exogenously, = 𝑛, where, ‘n’ is natural growth rate of population, equation (3) can be rewritten as
𝐿

𝑔 = (𝛼 + 𝛽). 𝑔 + (1 − 𝛼). 𝑛
𝛽.𝑛
or, 𝑔−𝑛 = 1−(𝛼+𝛽)

here, g-n shows the growth of per capita income. In absence of any spillover effects of technological
change, the constant returns to scale will be applicable and under such conditions since β=0, this will mean that
g-n will also be equal to zero which indicates that in absence of spillover effects, the economy will not
experience any growth in its per capital income and hence, the constant returns to scale will be applicable.
However, in Romer’s model all the factors viz. capital stock, labour and technology are assumed to be working
together whose productivities mutually influence each other, the value of β will always be positive. Hence, if β
> 0, the growth of per capita income will also be positive, therefore, g-n > 0. This is possible only due to the
spillover effects of the technology within as well as across production units.

Although, this model has provided an important breakthrough, yet its applicability to the
developing economies is questionable because many of the assumptions of this model find little validity in the
developing economies e.g. the assumption of single sector economy may be unrealistic for the dual economies
of the developing countries. Moreover, the developing countries largely face the problem of structural rigidities
55
which are hardly mentioned by Romer. The analysis of these rigidities is very important in the context of the
spillover effects of technology, research and development or any any type of knowledge across all production
sectors of the economy. Due to these structural bottlenecks many a times the developing economies are found
not to be using the full capacity of their available capital even though they fight with the problem of shortage
of

capital. Romer Model is silent about all such factors in the context of developing economies while
these very factors actually lead to slower growth of the per capita income in these economies even though they
are using the same technology as has been used by the developed economies. Romer’s model is silent about the
causes and effects of all such problems of the developing countries. Actually, the developing economies lack
sufficient incentives to invest in physical as well as human capital. This has a great effect on supply of savings,
capital formation and hence on growth of income. Besides, during the transition phase developing countries
also undergo the process of reallocation of resources which are not generally efficient ones, particularly during
the initial phase of the transition. This inefficient reallocation of resources at any point of time has medium as
well as long term effects upon the growth of income of the ecomnomy. But all these factors have been ignored
by Romer’s model. Hence, the developing economies find little guidance from this model.

Lucas Model

Lucas’ model of growth emphasises the importance of human capital in the growth the economy. He
states that it is difference in attainments of the human capital that has led to worldwide economic disparities.
Lucas states that the developed countries went through the process of industrial revolution long time back in
the history of these economies. The incentive to earn more profits has led to invest in creation of knowledge so
that the conditions of normal profits can be converted in to the long run capacity to earn supernormal profits.
However, this process was hardly understood by the developing economies. This is their misconceptions or
overindulgence in the idea of capital stock being the sole and most important determinant of economic growth
that has led to wrong strategies and hence they were not able to experience the same level of growth as
experienced by the economies which have experienced the epochs of industrial revolution. These
misconceptions or little understanding of the importance of human capital has led to lower investments in
human capital for a long time in these economies. Consequently, these economies have lagged much behind
the developed economies who are growing at faster rate leading to divergence across the poor and rich
economies of the world. Actually, the physical capital and human capital are not substitute to each other, they
are rather essential complements of each other. Therefore, a greater investment in human capital also leads to
greater productivity of the physical capital. In this framework Lucas model emphasises that the skilled workers
and the new technology are inseparable from each other. In order to measure the effect of human capital
accumulation on income, we can identify two separate components of total savings in an economy – these
savings can be used for increase in physical capital stock and/or these savings can also be used for enhancing
the level of human capital in the economy which will lead to an increase in productivity of labour as well as
capital in future time period. To further elaborate the growth model given by Lucas, let us first look at the basic
equation of this model. in this equation output is considered to be the function of physical as well as the human
capital stock.

𝑦 = 𝑘 𝛼 ℎ1−𝛼 ------------------------ (1)

56
 Here, ‘h’ stands for human capital and ‘k’ for physical capital. As discussed above, part of the savings
are spent in accumulation of physical capital and its part is spent on accumulation of human capital. These two
components of savings can be expressed in the following manner. Firstly, taking ‘s’ as the part of savings
being used for accumulation of physical capital i.e

𝑘(𝑡 + 1) − 𝑘(𝑡) = 𝑠𝑦(𝑡) ----------------------- (2)

Secondly, the proportion of savings spent on accumulation of human capital can be expressed as

ℎ(𝑡 + 1) − ℎ(𝑡) = 𝑞𝑦(𝑡) ------------------------ (3)

Thus, sy(t) and qy(t), respectively show the total amount of resources spent on accumulation of
physical and human capital. For self-sustained growth of the economy, ‘y’, ‘k’ and ‘h’ should be growing at
the same rate. The rate of economic growth actually depends upon the growth of investment in physical as well
as human capital. Therefore, it is important to find the ratio of investment in human capital to that of physical
capital. For this, we would have to find the growth of these two types of capital in an economy. a). Growth of
Physical Capital: The growth of physical capital can be derived from equation (2) by putting the value of y(t)
and also dividing both sides by k(t). The resultant equation can be written as
1−𝛼
𝑘(𝑡 + 1) − 𝑘(𝑡) 𝑠. [𝑘(𝑡)]𝛼 [ℎ(𝑡)]1−𝛼 ℎ(𝑡)
= = 𝑠. [ ]
𝑘(𝑡) 𝑘(𝑡) 𝑘

If we take h(t)/k(t)=r, this equation can be written as

𝑘(𝑡 + 1) − 𝑘(𝑡)
= 𝑠𝑟 1−𝛼
𝑘(𝑡)

Similarly, we can also derive the equation for growth of human capital

b). Growth of Human Capital:

ℎ(𝑡 + 1) − ℎ(𝑡)
= 𝑞𝑟 −𝛼
ℎ(𝑡)

Since, in the long period the growth of human capital as well as the physical capital are equal,
therefore,
𝑞
𝑠𝑟 1−𝛼 = 𝑞𝑟 −𝛼 𝑜𝑟 𝑟 =
𝑠
This ‘r’ can be used as long term growth rate and since in the long period, the growth of income,
physical capital and human capital are the same, therefore,

𝑦(𝑡 + 1) − 𝑦(𝑡)
= 𝑠𝑟 1−𝛼 = 𝑞𝑟 −𝛼
𝑦(𝑡)

= 𝑆 𝛼 𝑞1−𝛼

57
 Thus, the long term growth of the economy depends upon the rate of physical capital formation as well
as human capital formation. It is the human capital which compensates the fall in growth of output due to
applicability of diminishing returns on physical capital. The human capital investment, rather ensures
increasing returns by its internal as well as external positive and output stimulating effects. The internal and
external effects of human capital formation in any economy can be discussed as below

(i) Internal Effects of Human Capital: According to Lucas, the total time of a human being,
particularly a worker, can be divided in to two components – the time spent in production and the time spent in
accumulation of human capital. If we denote the proportion of time spent in production as µ(h), then the time
spent in accumulation of human capital will be 1- µ(h). In any economy, the size of the labour force as well as
its productivity per hour significantly influences the level of output. Therefore, instead of having merely the
size of the work force, Lucas has put forth the idea of ‘effective labour force’ which is shown as the product of
size of the labour force and the time spent on producing goods and services for a given level of human capital
∞

𝑁𝑒 = ∫ 𝜇(ℎ). 𝑁(ℎ). 𝑑ℎ
0

Here, N(h) is the size of labour force and Ne is the effective labour force. Thus, we can express
production as function of physical capital stock and effective labour force.

𝑌 = 𝑓(𝐾, 𝑁𝑒 )

The level of human capital would not only have a macro economic impact upon the aggregate output of
the economy but will also have accrue certain private benefits to the holders of the human capital as in a
competitive market economy, wages are paid according to the marginal product of workers. Since, the workers
with higher level of human capital are more productive, they would have higher earnings. Total wages in an
economy for a given level of human capital can be calculated as follows:

𝑇𝑜𝑡𝑎𝑙 𝑤𝑎𝑔𝑒𝑠 = 𝑓 ′ (𝐾, 𝑁𝑒 ). ℎ. 𝜇(ℎ)

Where, 𝑓 ′ (𝐾, 𝑁𝑒 ) shows the marginal productivity of labour

(ii) External Effects of Human Capital: Increase in level of human capital formation in any economy,
undoubtedly increase the level of productivity of a single worker but also have an overall effect upon the
average productivity of the economy as a whole. Even a single worker with higher human capital in a
production unit has huge ripple effects in the production unit. Same is true for the economy as a whole. But in
a perfectly competitive economy, it is generally assumed that human capital of an individual would not affect
the average level of human capital, yet its opposite is not true as average level of human capital in any
economy determines the minimum target to be achieved by average workers to ensure their employability.
Hence human capital investments and its attainments by the private individuals are largely determined by the
average level of human capital for the country as a whole. This is termed as external effect of human capital. In
order to know about the external effect of human capital, it is important to know about the average level of
human capital which can be calculated as below.

58
 ∞
∫0 ℎ. 𝑁(ℎ). 𝑑ℎ
ℎ𝛼 = ∞
∫0 𝑁(ℎ). 𝑑ℎ

Here, ha is average level of human capital in a country. Now, we can easily adjust our production
function by incorporating internal as well as external effect of human capital. First of all let us have production
as function of capital and effective labour force.
1−𝛽
𝑄 = 𝐴. 𝐾𝛽 (𝑡) 𝑁𝑒 (𝑡)

Putting the value of Ne, we get

𝑄 = 𝐴. 𝐾𝛽 (𝑡)[𝜇(𝑡). ℎ(𝑡). 𝑁(𝑡)]1−𝛽

Incorporating the external effects of human capital or the average level of human capital for the society
as a whole

𝑄 = 𝐴. 𝐾𝛽 (𝑡)[𝜇(𝑡). ℎ(𝑡). 𝑁(𝑡)]1−𝛽 [ℎ𝑎 (𝑡)]𝑟

Any increase in time for accumulation of human capital i.e. 1- will raise the individual as well as the
average level of human capital of the society which will have positive effects upon the level of output at an
increasing rate. Thus, the economy would grow at a faster rate due to applicability of increasing returns to
scale in the production sector. The change in human capital, which is the main force behind the faster growth
of the economy, can be measured as

ℎ̂(𝑡) = [ℎ𝑎 (𝑡)]𝑟 . 𝐺[1 − 𝜇(𝑡)],

here, G is the growth of human capital and it is always positive i.e. G > 0 but the existing level of
human capital or say the knowledge, which the society has attained so far, will have diminishing returns to
output, therefore,  < 0. In order to simplify the analysis if we simply assume  =1, then the equation showing
the change in human capital at any point of time can be shown as

ℎ̂(𝑡) = [ℎ𝑎 (𝑡)]. 𝐺[1 − 𝜇(𝑡)]

We can discuss here two extreme cases, one is when whole of the time is spent in accumulation of
human capital i.e. ()t = 0, and the second when whole of the time ()tis spent in production only i.e. 1- = 0 .
In the first case, the change in human capital will be

ℎ̂(𝑡)
=𝐺
[ℎ(𝑡)]𝑟

i.e. the economy can achieve the highest growth of human capital equal to G while in second case there
will be no change in growth of human capital and it would be zero. In absence of any change in human capital,
since diminishing returns to scale are applicable to the existing level of human capital, the economy will also
grow at diminishing rate for any change in its inputs. But in real life, the value of ()t or ()tthat of 1- varies
between 0 and 1 which shows that the economy moves on a continuous growth path and the rate of growth of
the economy will be higher for higher growth of human capital. This fact points towards the fact that the

59
applicability of diminishing returns can be postponed by increasing investment in human capital. Thus, the
economies with higher rate of growth of human capital experience a higher growth of income and the
economies with lower investment in human capital will experience the lower growth of income, leading to
divergence between the two types of the countries. Through this fact, Lucas pointed out that the the gap
between growth of rich and poor economies can be explained by the gap in investments in human capital in
these economies.

Finally, Lucas also differentiated between the optimum growth path and the equilibrium growth path.
By optimal path he means that the society wants to maximise its utility function by achieving the optimal level
of per capita income with an optimal combination of K(t), N(t), c(t), h(t) and ha(t). On the other hand,
equilibrium path means that there is simultaneous equilibrium in all firms and households as well the economy
as a whole. Assuming that ha(t) is exogenously determined and it is expected that each individual will follow
the same path so that the actual behaviour coincides with the expected behaviour and there is no gap between
demand and supply (i.e. AD = AS), for given physical and human capital stock. The solution will be achieved
in both the cases if h(t) = ha(t). Any divergence between the two will mean the divergence from the
equilibrium as well as the optimal growth path.

Like Romer’s Model, Lucas’ Model also gave little attention to the structural rigidities of the
developing economies. There is no doubt that these economies have low level of productivity due to low level
of human capital but at the same time these economies are also suffering from the problem of misallocation
and misutilisation of the available resources. These economies not only face the problem of skilled workers but
also underutilisation of existing human capital. Due to lack of opportunities and low returns to human capital,
there is little incentive to invest in human capital by private individuals. On the other hand, the skilled workers
have a greater tendency to migrate to other countries in search of better opportunities. These are the workers
which the economy needs the most and their emigrations means the drain of most essential and productive
resources. This loss of intellectual capital has a huge and long run adverse effect upon the economic growth of
the poor countries. These aspects are ignored by Lucas’ model of endogenous growth, yet there is no doubt that
the developing economies can learn a lot from this model that it is the higher level of human capital that can
ensure higher productivity of other resources. So, it must be attained as well as retained.

Summary

The neo-classical growth theories, though recognised the importance of technology but they took it as
exogenous factor and therefore they failed to explain why even in the long run the richer economies are
growing at a rate higher than the poor economies and why same level of investment and technology do not
produce similar results across the economies. Moreover, there is little evidence of convergence between the
two types of the economies as suggested by Solow’s framework. The answers to many of these questions can
be found in the endogenous growth models which state that the returns to capital and a given technology
mainly depend upon the inherent characteristics of the economy and therefore, the factors to growth are
endogenous rather than being exogenous. This is the spillover effect of the technology and the spread of human
capital that leads to increasing returns to scale even in the long run. Higher is the speed of these spillovers,
higher will be the rate of growth of the economy. Similarly, the economies which invest more in human capital
are also able to grow at a higher rate than the economies which spend less on the same. Yet, these models have
proposed one sector economy which finds little applicability in the developing economies which are dual in
character and face the problem of structural rigidities. In the developing economies, misallocation of the
60
resources is as big an issue as the availability of the same. Though the endogenous growth models are the
important breakthrough in the existing knowledge of growth economics but they hardly deal with the important
issues of the developing economies which hinder the process of growth even with increase in investment in
human capital as well as creation of technology.

AK MODEL — EXPLANATIONS OF CROSS COUNTRY DIFFERENTIALS IN ECONOMIC

GROWTH

Introduction

The neo-classical approaches to economic growth were largely considered to be unsatisfactory due to
several inherent flaws. These models view improvements in total factor productivity (technological progress)
to be the ultimate source of growth in output per worker, but they do not provide an explanation as to where
these improvements come from. In the language of economists, long-run growth is determined by something
that is exogenous in the model. Diminishing returns to the accumulation of capital, which plays a crucial role
in limiting growth in the neoclassical model, is an inevitable feature of an economy in which the other
determinants of aggregate output, namely technology and the employment of labour, are both given. However,
there is a class of model in which one of these other determinants is assumed to grow automatically in
proportion to capital, and in which the growth of this other determinant counteracts the effects of diminishing
returns, thus allowing output to grow in proportion to capital. These models are generally referred to as AK
models, because they result in a production function of the form Y = AK, with ‘A’ constant. The AK model is
actually considered the first version of endogenous growth theory. However, the earlier version of this model
go back to Harrod (1939) and Domar (1946) who assumed an aggregate production function with fixed
coefficients. Frankel (1962) developed the first AK model with substitutable factors and knowledge
externalities, with the purpose of reconciling the positive long-run growth result of HarrodDomar with the
factor substitutability and market clearing features of the neoclassical model. The Frankel model showed a
constant savings rate, whereas Romer (1986) developed an AK model with intertemporal consumer
maximization. The idea that productivity could increase as the result of learning-by-doing externalities, was
put forth by Arrow (1962). Then Lucas (1988) developed an AK model where the creation and transmission of
knowledge occurs through human capital accumulation. Similarly, we can cite a number of other models which
have followed the AK framework. Hence, it is important here to examine this approach and its contribution to
economic theory.

The Cross-Country Difference in Growth

There are large differences in per capita income across countries. Of total world income, 42 per cent
goes to those who make up the richest 10 per cent of the world’s population, while just 1 per cent goes to those
who make up the poorest 10 per cent (World Bank, World Development Indicators, 2014). This points towards
not only unequal distribution of world income across different countries but also differences in their growth
rates. The key sources of these differences can be numerous depending upon the national policies and
institutions. Hence, it is very important to understand how some countries can be so rich while some others are
so poor as the income differences have major welfare consequences. The differences in growth rate across
economies have actually widened the income inequalities. Acemoglu (2007) has indicated that even in the
historically brief postwar era, the world has witnessed tremendous differences in growth rates across countries
and these have ranged from negative growth rates to average rates as high as 10 per cent a year. During this
61
period some of the countries have grown at a faster pace, some at a slower rate and some stagnated after
growing for a short period. It is being believed that much of these differences in economic growth cannot be
wholly attributed to the post-war era alone as during this period, the “world income distribution” has been
more or less stable, with a slight tendency towards becoming more unequal. Further, the Maddison data has
suggested that much of the divergence took place during the 19th century and early 20th century. It is
important to observe that the process of rapid economic growth started in the 19th, or perhaps in the late 18th
century and then takes off in Western Europe, while many other parts of the world do not experience the same
sustained economic growth. The high levels of income today in some parts of the world are owed to this
process of sustained economic growth, and this process of differences in economic growth has also caused the
divergence among nations. This divergence took place at the same time as a number of countries in the world
started the process of modern and sustained economic growth. Therefore understanding modern economic
growth is not only interesting and important in its own right, but it also holds the key to understanding the
causes of cross-country differences in income per capita today. The endogenous growth theories largely owe
these differences or divergences in economic growth to the institutions, policies, technologies along with the
levels of investments and other transitional dynamics. These theories also pointed out the importance of
investment in human capita along with that of physical capital to explain these divergences. These theories also
point out that the technology differences across countries include both genuine differences in the techniques
and in the quality of machines used in production, but also differences in productive efficiency resulting from
differences in the organization of production, from differences in the way that markets are organized and from
potential market failures and how the human factors handle these technologies with effects on productive
efficiency. Hence, a detailed study of “technology”, physical capital and human capital as correlates of
economic growth is necessary to understand both the world-wide process of economic growth and cross-
country differences. It is important to examine the sources of income differences among countries that have
(free) access to the same set of technologies, but do not generate sustained long-run growth. A full analysis of
both cross-country income differences and the process of world economic growth requires models in which
technology choices and technological progress are endogenized. Hence, we can start with simple AK model.

Simple AK Model

As we have already discussed that the first version of endogenous growth theory was AK theory, which
did not make an explicit distinction between capital accumulation and technological progress. In effect it
lumped together the physical and human capital whose accumulation is studied by neoclassical theory with the
intellectual capital that is accumulated when innovations occur. An early version of AK theory was produced
by Frankel (1962), who argued that the aggregate production function can exhibit a constant or even increasing
marginal product of capital. This is because, when firms accumulate more capital, some of that increased
capital will be the intellectual capital that creates technological progress, and this technological progress will
offset the tendency for the marginal product of capital to diminish. In the special case where the marginal
product of capital is exactly constant, aggregate output Y is proportional to the aggregate stock of capital K:

𝑌 = 𝐴𝐾

where A is a positive constant that reflects the level of technology and ‘K’ here is taken in a broader
sense as it includes physical as well as human capital. This model shows constant marginal product to capital
(as 𝑀𝑃𝑘 = 𝑑𝑌/𝑑𝐾 = 𝐴) indicating that long run growth is possible. Thus, AK model is a simple way of

62
illustrating endogenous growth. Assuming a closed economy, the savings are equal to investment under
conditions of full employment.

Since savings are the function f income and capital depreciates at a constant rate i.e. ‘δ’ the change in
capital stock can be traced through following equations.

𝐼 = 𝑆 = 𝑠. 𝑌 = 𝑠. 𝐴𝐾

and, since capital depreciates at a constant rate, the change in capital stock i.e. 𝐾̇ can be expressed as 𝐾̇
= 𝑠. 𝑌 − 𝛿.𝐾. This change in capital stock can also be represented by a diagram given below

In this figure Y-axis show output per worker while the X-axis show the capital stock. The line Y=AK
having a constant slope shows the constant marginal productivity of capital; the line S=s.Y is the gross
investment line while the line δK shows the depreciation line or the total replacement investment. The
difference between the gross investment line and the replacement line i.e. area between S=s.Y line and δK line
shows net investment in the economy which is positive and increasing.

The growth of capital stock can be found by dividing both sides of the equation showing change in
capital stock with ‘K’, we get

𝑌̇
Further, assuming that s.A > δ, the growth of capital stock as well as growth of output i.e.𝑌 > 0,
showing that the economy will be ever increasing as compared to the Solow model which shows that the
income increases at a declining rate before stagnating for a given technology. We can easily compare how the
63
paths of per capita income would differ in case of Solow model of exogenous growth and the AK model of
endogenous growth. This comparison has been shown through figure 2

Source: Miguel Lebre de Freitas, Introduction to Economic Growth.

Figure 2 compares the impact of rate of change in savings upon the growth of income. The top part of
the diagram shows the levels of income and the bottom part shows the growth rate of the same. In the upper
part, we can see that a once for all increase in saving rate in t0 time period leads to an ever growing income
curve (shown as ln y) in case of AK model while in case of Solow model, the income increases initially but
ultimately reaches at the same level after t1. This can be observed through the angle ‘γ’. In case of Solow type
growth path, as savings increase or say, due to exogenous change in technology in t0 time period, the income
curve immediately and its slope rises as we can see that the size of angle ‘γ’ increases from γ0 to γ1 but after t1
time period, it again comes back to the previous level i.e. γ0. However, in case of AK type growth the increase
in income is forever, shown by an ever increasing curve and once for all increase in the size of angle γ0 to γ1.
The growth path can better be elaborated through growth rate of income in the lower segment of the diagram.
We can see that in t0 time period, the growth rate of income increases immediately from γ0 to γ1 but the Solow
type growth path shows that it comes back to γ0 level while the AK growth path shows that growth of income
is constant at stays at γ1 level even in the long run.

AK Model with Human Factor

In its more realistic form, we can also add labour as an input along with capital. In this context, first of
all, we can discuss Arrow’s model with knowledge spillovers. In this model, the production function for final
output can be written as

𝑌 = 𝐵.𝐾 𝛼𝐿 1−𝛼 (1)

which is a Cobb-Douglas type production function showing constant returns to scale with inputs K and
L. In a model with technology and population growth as exogenous factors, the population, equal to labour
input L, can be normalized to one and the individual firm takes total factor productivity B as given. However,

64
we suppose that B is in fact endogenously determined. Specifically, the accumulation of capital generates new
knowledge about production in the economy as a whole. In particular, we assume that

𝐵 = 𝐴𝐾 1−𝛼 (2)

where, A is constant and is greater than zero i.e. A > 0 That is, an incidental by-product of capital
accumulation by firms in the economy is the improvement of the technology that firms use to produce.
Technological progress, modelled as a by-product of capital accumulation, is external to the firm. Combining
the two preceding equations gives

𝑌 = 𝐴.𝐾. 𝐿 1−𝛼 (3)

This is exactly the AK model above, noting that L = 1. However, in further formulation of the AK
model, we can include human capital as a separate variable having a positive effect upon the level of output.
Thus, more skilled labour force will be assumed to produce more output than an unskilled individual, and the
total stock of such “skills” is called human capital. Crucially, human capital can be accumulated through
education. Thus, both types of capital can be accumulated—this turns out to imply that the model has similar
properties to the AK model. In this perspective, we can have a production function of the following type:

𝑌𝑡 = 𝐴𝑡 .𝐾𝑡 𝛼𝐻𝑡 1−𝛼 (4)

where Kt is physical capital and Ht is human capital. So, growth is determined by

𝑌̂𝑡 𝐴̇ 𝐾̇ 𝐻̇
= 𝐴𝑡 + 𝛼 𝐾𝑡 + (1 − 𝛼) 𝐻𝑡 (5)
𝑌𝑡 𝑡 𝑡 𝑡

Assuming that like physical capital, human capital also depreciates for given attainments. This can be
understood in this way that if a person does not updates its knowledge, the knowledge accumulated so far
depreciates in a dynamic economy. For simplifying the analysis, let us assume that both the physical as well as
human capital depreciates at the same rate, then we can easily derive the equations for accumulation of each
type of capital stock. This simplification is expected not to disturb the relevant conclusions of the model.
Hence, the change in physical capital stock is defined as

𝐾̇ 𝑡 = 𝑠𝑘𝑌 – 𝛿𝐾 (6)

and the change in human capital stock can be defined as

𝐻̇ 𝑡 = 𝑠𝐻𝑌 − 𝛿𝐻 (7)

Similarly, we can also define the capital output ratios for both the physical and human capital

The capital output ratio for physical capital is

𝐾
𝑥𝑘 =
𝑌
while for human capital, it is

𝐻
𝑥𝐻 =
𝑌
65
For finding the growth rate of physical as well as human capital, we can divide the equations (6) and (7),
respectively with ‘K’ and ‘H’ and substitute the above values of respective capital output ratios, we have
𝐾̇ 𝑠
= 𝑥𝑘 − 𝛿 (8)
𝐾 𝑘

and
𝐻̇ 𝑠
= 𝑥𝐻 − 𝛿 (9)
𝐻 𝐻

In AK model, for a steady growth path, the growth for physical and human capital stock should be the same i.e.
̂
𝐾 𝑠𝐾 𝑠 ̂
𝐻
= − 𝛿 = 𝑥𝐻 − 𝛿 = 𝐻 (10)
𝐾 𝑥𝐾 𝐻

This implies that

𝑠𝐾 𝑠 𝑥 𝑠 𝐾 𝑠
= 𝑥𝐻 ⇒ 𝑥𝐾 = 𝑠𝐾 ⇒ 𝐻 = 𝑠𝐾 (11)
𝑥𝑘 𝐾 𝐻 𝐻 𝐻

So, along the steady growth path, the stock of human capital can be written as
𝑠
𝐻 = 𝑠𝐻 𝐾 (12)
𝐾

This value of ‘H’ can be put in equation (4), we get

𝑠𝐻 1−𝛼
𝑌 = 𝐴𝐾 𝛼 ( 𝐾)
𝑠𝐾

Or,

𝑠 1−𝛼
𝑌 = 𝐴. 𝐾 (𝑠𝐻 ) (13)
𝐾

Again, under the steady growth path, growth of income is equal to the growth of capital stock

𝑌̇ 𝐾̇
=
𝑌 𝐾
𝐾̇
From equation (8) and putting the value of 𝑥𝑘 = , we get
𝐾

𝑌̇ 𝑠𝐾 𝑌 𝑠𝐻 1−𝛼
= − 𝛿 = 𝐴. 𝑠𝐾 ( ) −𝛿
𝑌 𝐾 𝑠𝐾

Hence,
𝑌̇
= 𝐴. (𝑠𝐾 )𝛼 (𝑠𝐻 )1−𝛼 − 𝛿 (14)
𝑌

Thus, allowing for both type of inputs- physical as well as human capital, which are continuously
accumulated produce same results as that of the AK model as equation (14) is another form of growth of
66
income in AK model with capital input only. In this case, the steady growth rate is 𝐴. (𝑠𝐾)𝛼 (𝑠𝐻)1−𝛼 − 𝛿 instead
of AS – δ as in simple AK model. Here, the simple saving rate has been replaced with a geometric average of
the two saving rates in the two factor model while leaving the broader implications unchanged. Lucas on the
other hand tried to incorporate the role of human capital in terms of effective labour force and an attempt has
also been made to measure the human capital accumulation in terms of allocation of time between production
time and time spent in human capital accumulation. Thus the effect of human capital accumulation can be
observed in the standard AK model. Although, we have already discussed the Lucas model in the lesson on
endogenous growth models, yet it is important here to look at it briefly in AK framework. For this purpose, we
now consider a simple endogenous growth model with human capital accumulation in which Y is a function in
physical capital K and "effective labour" h. L, where ‘h’ is level of human capital per person and ‘L’ is the size
of the labour force.

𝑌 = 𝐾 𝛼 (ℎ. 𝐿)1−𝛼

Lucas assumes that human capital per person evolves according to

ℎ̇ = (1 − 𝑢)ℎ

where u is time spend working and (1 – u) is time spend accumulating skills. The growth rate of h is given by

ℎ̇
= 1−𝑢
ℎ
Since h enters the production function like labour-augmenting technological progress in the Solow model, we
can conclude that the long-run growth rate of output is equal to the growth of human capital stock which is
𝑌̇ ℎ̇ 𝑌̇
equal to the growth of physical capital stock on steady state equilibrium path. Hence, 𝑌 = ℎ 𝑜𝑟 𝑌 = 1 − 𝑢

This implies that every policy measure which affects u has an impact on the long-run growth rate.

Thus, we have observed that AK model gives a new framework for the long run growth of the
economies. However, there are still some reasons to doubt the predictions about long-run growth generated by
this class of models. The first line of criticism is related with the non-accumulable factors. In the real world,
there are factors of production that are in fixed supply, such as land, or that cannot simply be accumulated
indefinitely such as energy. Remember that the AK model results are of a knife-edge variety: Any move away
from all factors being accumulable, and we move back to the Solow model results. Moreover, similar treatment
to all type of human capital is also criticised by many as they say that the strict parallel between human capital
and physical capital in the model just described is probably not completely accurate. For instance, not all
expenditures on education will produce the same effect on output. The marginal boost to aggregate output of
primary teaching is altogether different to that of higher education; training the nonskilled informally and the
formal training of the professional also differ in their marginal returns. By clubbing all these different types of
human capital together hardly proposes an effective policy suggestion for countries with a varied structure of
human capital. Another source of the difficulties faced by the AK model is that it does not make an explicit
distinction between capital accumulation and technological progress. In effect it just lumps together the
physical and human capital. Lastly, it also criticised for not giving any explanation for any possibility of
convergence among the economies.

67
Summary

Now let us summarise what we have learnt in this module:

• AK model is considered the simplest version of endogenous growth. Although, we can observe many
neo-classical approaches in the AK framework such as that of Harrod and Domar who assume an
aggregate production function with fixed coefficients but the important implication of modern AK type
production function is that it can explain long-run growth using the same basic assumptions as the
neoclassical model but adding knowledge externalities among firms that accumulate physical capital.
• The idea that productivity could increase as the result of learning-by-doing externalities, was most
forcefully pushed forward by Arrow and then Lucas developed an AK model where the creation and
transmission of knowledge occurs through human capital accumulation.
• All these models clearly state that the economies can experience long term growth or the increasing
returns to scale through knowledge spillovers and since the developed countries have a higher stock of
human capital as compared to the developing ones, the former grow at a faster pace than the latter one
and the economies of two types of economies go on diverging from each other.
• However, like all the endogenous growth models, the AK model also does not address the structural
rigidities of the developing countries which hinder the process of creation as well as accumulation of
human capital which is an important factor for ensuring sustained long run growth of the economies.

68