CHAPTER

25 NEW (ENDOGENOUS)
GROWTH MODEL
(ROMER MODEL)
1. INTRODUCTION increasing returns to scale. This in turn allows
At various times in the history of thought, investment in knowledge capital to persist
economists have stressed increasing returns as indefinitely and to sustain long-run growth in
an endogenous explanation of economic growth. Per cap1ta income.
Adam Smith did so in emphasizing that growth Comparison between New Growth theory and
in productivity was due to the division of labour, Neoclassical theory
which depends upon the extent of the market. It would be useful to understand the
Alfred Marshall also emphasized that the role
of "nature" in production may be subject to difference between new (endogenous) growth
diminishing returns, but the role of "man" is theory and the neoclassical theory. One way to
subject to increasing returns. J.M. Clark also explain the contrast between them is to recognise
observed that "knowledge" is the only instrument that many endogenous growth theories can be
of production that is not subject to diminishing expressed by the simple equation Y - AK. In
returns"} Allyn Young also related economic this equation A represents any factor that affects
progres to increasing returns ... as a result of technology,K includes both physical and human
progressive division and specialisation among capital. But notice that there are no diminishing
industries and the use of round about methods returns to capital in this formula and possibility
of production.? exists that investments in K (physical and human
2. NEW GROWTH THEORY capital) could generate external economies
sufficient to offset the diminishing returns. The
Robert Solow's neo-classical growth model net result is sustained long-term grow th-an
explained that diminishing returns were outcome prohibited by traditional neoclassical
applicable to capital and labour separately and growth theory.
constant returns to both imputs jointly and
treated technical progress as a residual?. The Again new growth theory highlights the
new growth theory examines production importance of savings and human capital
functions that show increasing returns because investments leads to
for achieving rapid growth, but it
several implications that are in direct
of specialisation and investment in
capital. Technical progress and human capital knowledge conflict with traditional theory.
formation are endogenised within general () There is no mechanism or force which can
equilibrium models of growth. New knowledge lead to the equilibration of growth rates across
is generated by investment in
The technical progress research sector. countries. National growth rates differ across
residual is accounted for countries, depending upon their savings rates
by endogeneous human capital formation, With and technology levels.
knowledge being treated as
over benefits to other firmspublic good, spill (ii) There is no tendency for convergence per
may then allow capita income levels in poor countries and
a8gregate investment in knowledge to exhibit of rich countries. The absence of hoe
convergence
1. J. Maurice Clark, Studies in
the Economics of Overhead Costs,
2. Allyn A. Young, 1923, p. 120.
"Increasing Returns and Economic
J. Kobert M. Solow, "A Contribution to the Theory of Progress, "Economic Journal, December 1928, pp. 527-42.
1956, pp. 65-94. Economic Growth". Ouarterly lournal of Economics, ebruay

264
 ENDOGENOUS GROWTH MODEL (Romer Model) 265

leads to a situation where there is a possibility growth model. This model addresses
of greater income gap between poor and wealthy technological spillovers that may be present in
Countries. the process of industrialisation. Thus it is not
( ) The interesting aspect of the endogenous only the seminal model of endogenous growth,
growth models is that they help explain erratic but one of particular relevance for developing
countries. We use a simplified version of
international flow of capital that widens the Romer's model that keeps his main innovation
economic disparities between developed and in modelling technology spillovers-without
developing countries. The potentially high rates presenting unnecessary details of saving
of return on investment offered by developing determination and other general equilibrium
economies with low capital-labour ratios are issues.
eroded by the low level of complementary
investments (i.e., investment in education, The model begins by assuming that growth
process derives from the firm or industry level.
infrastructure, research and development etc.) Each industry, individually produces with
(iv) Unlike traditional neoclassical theory cOnstant returns to scale, so the model is
(Solow model), new growth models explain consistent with perfect competition and upto this
point it matches with assumptions of Solow
technological change as an endogernous outcome
of public and private investments in human model. But Romer departs from Solow by
capital and knowledge-intensive industries. assuming that economy wide capital stock K
positively affects output at the industry level so
Thus, in contrast to neoclassical grow th that there can be increasing returns to scale (IRS)
theories, endogenous growth models suggest an at the economy level.
active role of public policy in promoting direct The aggregate production function of
and indirect investments in human cap1tal Romer's model can be expressed as under :
formation, encouraging foreign private Y = AKa+ß.;l-a
investment in knowledge-in tensive industries
such as computer software and tele Here Y, K and L respectively represent output,
communications. capital and labour. To make endogenous growth
stand out clearly, we assume that Ais constant
3. THE ROMER MODEL rather than rising over time, that is, we assume
for now that there is no technological progress.
Models of endogenous growth bear some With some knowledge of differential calculus, it
structural resemblance to their neo-classical
can be shown that the resulting growth rate for
counterparts, but they differ considerably in their per capita income in the economy would be*
underlying assumptions and the conclusions
drawn. The most significant theoretical 8-n = pertapta
differences stem from discarding the neo classical 1-(a+B)
assumption of díminishçng marginal returns to growth rate and nis the
capital investments, permitting increasing where g is the output Without spillovers, as
returns to scale in aggregate production. By population growth rate.
assuming that public and private investments in the Solow model so per
with constant returns to
capita growth rate would
in human capital generate external economies Scale, B = 0 and
(without technological progress).
and productivity improvements that offset the be zero
natural tendency for diminishing returns, Romer assumes, however, that taking the
endogenous growth theory seeks to explain the three factors together including capital
existence of increasing returns to scale and externality : B> 0; thus g -n>0 and Y/L is
divergent long-run growth patterns among growing. Now, we have an endogenous growth,
countries. depending on the level of savings and investment
To illustrate the endogenous growth undertaken in the model, not driven exogenously
approach, we explain the Romer endogenous by increases in productivity. The interesting

The deriva tions of this equation is given on page 266

266 ENDOGENOUS GROWTH MODEL (Romer Model)

feature of Romer model is that with an K'

investment (or technology) spillover, the model constant and Y K -8 (constant growth rate)
avoids diminishing returns to capital.
and =n which is also constant.
*Production function is given by the equation L
Y =A Ka+ß. L-a ...(1)
By Chain rule method
Substituting the values of YY K8
K
and
aY OY OK + OY oL
Y' = ..(2)
dt OK at OL ot nin Eq. (3), we get
L
-and are partial differenti als
OK OL 8 =(a+B)g +(1-«)n
of Ywith respect toK and L 8-(a +B)g =(1-)n

To get the values of OK and OL, We

g1-(a +B)] =(1-a)n
differentiate partially the production function 1-a
OY
1-(a -B)
OK
= A( a+B)-K a+ß-1[l-a Subtracting n from both sides
OY 1-a
=A·Ka+ß (1- a)Ll-a-1 8-n n-n
OL 1-(a +B)
- A-Ka+ß -(1-a)La 1-a
OY
Substituting the values of and
OK OL in
Eq. (2), we get |1-a-1+a +ß
dY
1-(a +B)
Y' =
dt
8-1 =
= A(a +B) Ka+ß-l. !-a. K+ A-K u+ß-( 1-a) L-a.L' 1-(a +B)
This is the result which we require for the
OK
= K, discussion.
4. CONVERGENCE
=A·K *B .!"(a +p)-
K-K'+(1-a)l:·L' Before we take up the issue of convergence,
it should be clearly understood that Romer's
theory is not the only one called as 'new' theory,
but the theories propounded by Lucas, Arrow
and others having the same line of approach,
are also part of endogenous growth model. In
order to test the convergence of the 'new' growth
theory, we are to see whether or not poor
Dividing both sides by Y, we have:
countries do grow faster than the rich ones. In
Y other words, we are to see whether there is an
Y -(a +B) K (-a): ..(3) inverse relation between the growth of output
(or output per head) and the initial level of per
Y K' L' capita income. If there is an inverse relationship
For a steady growth y''T are all between the two, this would support the neo
 267
ENDOGENOUS GROWTH MODEL (Romer Model)
factors
classical theory. If there is not, this would convergence, holding constant all other
the growth of per capita incomter
Support the new growth theory's assertion that that influence
marginal product of capital does not decline. ratio including population growth (p), investmnent
following producthvy
(1/Y) and variables that affect(ED),
1o assesS this proposition, we use the education research
labour, for example
linear equation : of
development expenditure (R+D), trade (T)
and variables such as
8i = a + b, (PCY); ...(1)
and even
non-economic
stability (PS) measured by the numbeT
where 8i is the average growth of output per political and coups. To include
these
head of a country (i) over a number of years. a of revolutions
is constant and b, is the regression coefficient of be factors and other variables, the equation () can
the regression equation (1). PCY, is the initial expended as under :
bi (1/Y); + bË (ED);
level of per capita income of country (i). If 8i =a + b; (PCY); + b (p); + b;(PS);t... ...(2)
happens to be a significantly negative, then it + b5 (R+D); + b, (T); + b, what
would be a case of unconditional convergence or asked is
Now, the question to be
beta (B) convergence as it is called in the literature. happens to the sign of the initial per capita
implies that poor variables are
1nis type of convergence the rich. income variable (PCY) when other
COuntries are growing faster than social
sign turns
introduced into the equation ? If the made for
without allowing for any other economic, negative (b < 0) when allowance is
the countries. The would
or political differences between connection, have these and other factors,
this surely
various studies made in this model : that is,
of unconditional represent the case of neo-clasical
not able to find the evidence were not for
there wwould be convergence if it countries in
The es timate of bË is not rich and poor
convergence.
negative ; in fact it is invariably differences between important variables in the
significantly all these and other
positive, indicating divergence.4 New' growth theory would be
this is growth process. research and
conclusion that education,
Before arriving at the theory, it should. Supported by factors like and so on. These
neo-classical expenditure
a rejection ofremembered that the neo-classical developmentkeep the marginal product of capital
however, be factors help
prediction of convergence assumes that the from falling, producing actual divergence in the
population growth, world economy. An empirical illustration of
savings or investment ratio, that affect the what we have been talking about is given in
technology and all other factorssame across
productivity of labour are the Table 25.1.
assumptions are not true,
these
countries. Since
the presumption of The statistics given in Table 25.1 have been
there can never be over the period
unconditional convergence (even if there are compiled by taking 117 countries
diminishing returns
to capital), only conditional of 1960-86 (Pritchelt 1997). This table shows the
Table 25.1*

Explaining Growth of per Capita GDP, 1960-88

Unconditional Conditional The richer
divergence divergence accumulate faster
Average growth Average growth Investment Primary school
of GDP of GDP level enrolment
per capita per capita
Effect of (1) (2) (3) (4)
0.40 0.32 4.43 14.57
Initial level of GDP per capita
relative to leader
investment 0.07
Average level of school 0.03
Average enrolment in primary
'Divergence, Big Time', Journal of Economic Perspectives, 11, 1997.
"Source:L. Pritchett,
sduanced theoretical discuSsion of convergence 1$sues, See the Symposium in Economic lournol luly
1996.
Thirlwall, Gronoth and Development with Speciml Keerence to Developing Economies, Eighth Edition, 20o6
p. 155.
 268 ENDOGENOUS GROWTH MODEL (Romer Model)
Inefficiencies arising out duc to poor
rate ot growth of per capita income on the initial infrastructure, inadequate institutional structures
level of per capita income relative to the leading
country. It also shows no evidence of and imperfect capital and goods market impinge
on the flow of growth prOcess. The endogenous
unconditionalconvergence because the coefficient
growth theoryoverlooks these influential factore
of 0.4 is positive. It is interesting to note, however,
that when differences in investment and as such its applicability for the study of economi.
schooling between the countries are allowed íor,development is limited especially when count
to country comparisons are involved.
the coefficient becomes negative (-0.32) indicating
conditional convergence. The fact is that rich (3) Allocational inefficiencies : This theory
countries are able to save and invest more, and fails to explain low rates of factory canari
do devote resources to education whnich utilisation in low-income countries where canital
perpetuate their growth advantage. Columns is scarce. In fact poor incentive structures mav
and (4) in the above table show astrong be responsible tor sluggish GNP growth rate
nositive relation between the initial level of Per savings and human capital accumulation
canita income and the investment level on the Allocational inefficiencies are common in LDC
one hand, and primary school enrolment on the that are undergoing transition from traditional
other. Poor countries tend to grow more slowly to commercial markets.
than the rich ones inspite of the potential (4) Neglect of impact on short-run growth :
advantages conferred by backwardness.
Another shortcoming of endogenous growth
The gist of the entire discussion is that non model is that it ignores the impact of various
diminishing returns to capital or the constancy factors on short and medium-term growth. It
of the capital-output ratio, lies at the heart of over-emphasises on the determinants of long run
'new' growth theory, as propounded by Lucas growth rates. In this sense this theory provides
and Romar who emphasized externalities to a partial explanation of growth process.
education and research. Moreover, the empirical studies of endogenous
growth theories have offered only limited
5. CRITICISM support.
The new growth theory or endogenous (5) Closed economy : Another serious
growth model is replete with interesting features weakness of new' growth theory is that it is
like discarding the neo-classical assumption of based on the assumption of closed economy.
diminishing returns to capital investment and Many of the 'new' growth models are closed
focussing on the role of public and private economy models, and there are no demand
investment for generating external economies. constraints. It is difficult to imagine how growth
However, this model is criticised on following rate differences between the countries can be
grounds. explained without reference to trade, and without
(1) Inappropriate assumptions : The new reference to the balance of payments position of
growth theory is based on a number of traditional countries, which in most developing economies
constitutes the maior constraint on the growth
neo-classical assumptions that are often of demand and output. Where a trade
inappropriate for LDCs. For it assumes is included in the model equation, as variable
that there is asingle sector ofexample,
production or that done in
all sectors are symmetrical. This does not permit equation (2), it is invariably insignificant or loses
its significance when combined with other
the crucial growth-generating reallocation of
labour and capital that are transformed during variables.
the process of structural change. (6) Growth is not asmooth process : Amore
(2) Poor infrastructure : Another weakness who argues issue is raised by Pritchett (2000)°
fundamental
of this model is the existence of poor growth of that it is difficult to character1se he
infrastrúcture because of which economicgrowth time trendmany developing countries by snge
because grow th is very volate.
in developing countries is frequently impeded. Periods of rapid growth are often followed by
6. L. Pritchett, Understanding Patterns of Growth : Searching for Hills among Plateaus, Mountains and Plains,
World Bank Economic Review, May, 2000.
 269
ENDOGENOUS GROWTH MODEL (Romer Model)
silver lining
theory, it has a implication
steep declines. Rapid and slow growth are for the 'new' growth A policy
the most part transitory. Very few countries see for developing countries. promote
governments can in the
their success or failure persist decade to decade. of this model is that incentivesto agents
providing capital-intensive
Microeconomic instability would be one major growth by human
developing
explanation of volatility. Monetary, fiscal and knowledge-producing
model has a lesson for emphasis
exchangerate policies affect demand, investment sectors. This place more
and terms of trade. 'New' growth theory fails countries that they should
more than on
physical
capital-even benefits from
to address these issues. on human countries can draw
capital. These with an open
exchange of ideas that come economy.
Conclusion the the global
economy integrated into
Despite various limitations associated with

QUESTIONS
of Romer model of endogenous growth. approach?
1. Discuss the strengths and weaknessesapproach differ from the traditional(Neo classical or Solow) developing
(new) growth confronting the
2. How does endogenous Romer model in the context of development problems
Discuss the relevance of
3. countries.
'new' (endogenous) growth theory.
4. Outline the essential propositions of
SELECT REFERENCES
(July1988).
Economic Development", Journal of MonetaryEconomics, 22Perspectives,
Mechanics of Economic
1. Lucas, Robert B., "On the Intellectual Appeal and Empirical Shortcomings, Journal of
H., 'Endogenous Growth:
2. Pack, Economy, October, 1986.
Winter 1994. Journal of Political
Returns and Long run Growth,' 1994.
3. Romer, P.M. 'Increasing Endogenous Growth', Journal of Economic Perspectives, Winter,Development Studies,
4. Romer, P.M.,"The Origins of
Theory and Development
Economics : A Survey', Journal of
5. Ruttan, V., 'New Growth
December, 1998. 2011.
of Development, Ninth Ediion,
6. Thirlwall, A.P., Economics
Stephen C. Smith, Economic Development, Tenth Edition, 2009.
7. Todaro,Michael P. and