Economics of Developing Countries: The Lewis Model

Introduction

Many scholars have offered answers to why civilisations rise and fall and hat forces determine economic growth, but up to now there is no single universally accepted explanation. Some people have stopped searching for explanations to how economic growth takes place. They look at each case of a development process as unique since no two countries are exactly the same, just as no two individuals are exactly alike. But why do we need to try to understand economic growth? A developing country would need such an understanding to help in each effort to develop workable ways or strategies to achieved economic prosperity.

The study of income distribution has been a mainstay in the study of economics from the time of the physiocrats to the present day. Adam Smith’s Wealth of Nations, written shortly after the peak of the physiocrats, became the reference point, as it did for many branches of economics, for a theoretical underpinning of studies related to income distribution. Why is research on income distribution of importance to economists? It is of course interesting, in its own right as part of a general description of an economy. More importantly though are the implications that it may have. There are four important implications that we can immediately identify.

The first of these is that the level of income of an individual is one of the indicators of the well being of the individual. Given the average income of a country, the distribution of income provides some indication of how well off are different sectors of the population. The second implication is related to the growth of the economy. Growth is dependent upon investment. Whether investment occurs and at what rate, depends to some extent on who controls the available funds for investment.

Studies of income distribution therefore have often attempted to identify who controls these funds. The third relationship is an extension of the second. The distribution of income helps to determine who controls savings and investment, and it is also related to the distribution of wealth. Most social scientists agree that there is a causal relationship between wealth and power.

If this is correct then a study of the distribution of income may not only shed light on the process of growth, but also on the pattern of development that a particular economy may undergo. Different income distributions will produce different demands for goods and services and hence generate different patterns of investment. Different patterns of growth result from these investment decisions. In less industrialised countries, where the rate and pattern of growth is of great importance to policy makers, examination of income distribution may give an indication of how the economy is likely to develop.

The last general implication is the relationship between income and demographic change. The supply of labour is an important determinant of the evolution of an economy. Understanding the relationship between the supply of labour and income distribution is therefore important in understanding the development of an economy.

There are two parts to understanding this relationship. In the short term the level of income accruing to different sectors of the population has an impact on an individual’s decision on whether to participate in the labour force. In the long term changes in income have some effect on decisions that affect family size. Understanding how family size and therefore the size of the total population change gives us an indication of long-term labour supply.

Theories of Economic Growth

Theories of economic growth may be classified into three broad groups. One group consists of theories viewing economic growth as a natural and inevitable process. The second group of theories explain economic development as a rational process brought about when men respond to opportunities in the environment so as to promote their own self-interest or material welfare. The last group of theories views economic development as a result of seemingly economically irrational yet psychological and sociologically satisfying activities. (Bourguignon and Morrisson 1998)

Kelly’s model falls on the second thought. If we could recall the recent development in the so-called developing countries can be viewed within the perspectives of the Lewis Model after fifty-four years. (Fields 2004) The most recent is the ongoing “Free Trade” in the guise of GATT WTO or General Agreement on Tariff and Trade. Every member of this trade organization can freely exchange and engage in trade between member countries without or lower tariff. This would mean exportation of labours and capital can be transferred between countries to a lower labour cost. Lewis (1954) model stressed that a surplus of labour will result in lower wages thus the capitalist sector will increase their capital for reinvestment.

On the other hand if the labour is scarce then wages will increase in so doing the capitalist incur loses or less profit. Therefore the best position is which the subsistence sector and the capitalist are in mutual position when labour meet the needs of the capitalist and labour are paid equitably. In that sense a balance is achieve, the capitalist gains while the labour sector are satisfied. Definitive assessment of the Lewis Model can be reckoned to the development of the economy specifically among the under develop countries.

The Classical School and Income Distribution

Smith’s Wealth of Nations became the reference point for studies on income distribution. Smith’s views developed into what we now call the classical school. These views are not only of historic interest. They arouse our attention because of the influence they have had and continue to have in the study of developing countries and especially in the study of income distribution in developing countries.

One of the central concerns of the classical school was understanding how economies grow. It was from their study of growth that we can infer their views on income distribution. For the classical school growth was dependent on the accumulation of surplus. Surplus was defined as the return from production over and above the wage bill and the replacement of capital. How much investment took place was therefore intimately tied to the amount of surplus. This view of investment was based on an agricultural model of production (Weldon, 1983: 19). Under such a model surplus is what is left after the farmer or landowner has put aside what he or she needs to survive.

While the amount of surplus determined the limit of investment, it did not determine the actual amount. This was determined by a behavioural assumption about the owner of capital or land. Under this scheme the amount of investment depended on the thriftiness of the owner of capital (Smith 1976: 337). In Smith’s words:

“Parsimony and not industry is the immediate cause of the increase of capital. Industry indeed, provides the subject which parsimony accumulates. But whatever industry might acquire, if parsimony did not save and store up, the capital would never be greater.”

By assumption the only people who could save were owners of capital, who had sufficiently large incomes to do so. While landlords may have had enough income they were more inclined to employ retainers and engage in luxury spending with any money over and above subsistence that they may have had. Given this assumption growth took place in the sectors of the economy that were dominated by capitalists.

Smith and his counterparts modelled the economy as one having two sectors. A traditional mostly agricultural sector and a modern industrial sector. This view of the economy was influenced heavily by the time in which they lived. At that time the industrial revolution had just begun and large sections of the population were still involved in agriculture or traditional crafts. The modern sector was where the capitalists were to be found. Given the behavioural assumption about investment, this sector was therefore the engine of growth. Since all investment took place here, the amount of income accruing to this sector grew steadily. (Kanbur and McIntosh, 1989).

The traditional sector acted as a reservoir of labour for the modern sector. Within the analytical framework used by Smith its most important role was in acting as an anchor for wages. The average income derived from the traditional sector acted as a reservation wage for workers in the modem sector. If capitalists offered a lower wage than this they would attract no labour as it was assumed that workers could return to the traditional sector. Because of the ability of the worker to re-enter the traditional sector, the concept of unemployment as we know it did not exist. (Bourguignon and Morrisson 1998)

Smith’s and his follower’s aim was to understand how surplus was produced, who controlled it and how it was disposed. Their policy objective was to develop policies that increased the size of surplus and its accumulation. Karl Marx, writing nearly a century after Smith, had a different purpose. Smith wrote early in the industrial revolution at a time of great promise.

Marx wrote when the industrial revolution had become entrenched and its darker side was beginning to be exposed. Marx’s objective was not only to understand how surplus was produced and accumulated but also to lay bare the suffering undergone by a large section of the population that resulted from the accumulation of capital. In addition to examining the process of production and accumulation of surplus Marx also explored the economic conditions that resulted in social change. (Kanbur and McIntosh, 1989).

Dualism and Development Theory

Development theory has been dominated by views that characterise developing nations as dual economies. This characterisation, arising from the classical model examined above, depicts the economy as composed of two distinct sectors: a rural sector also known as the agricultural or traditional sector and an urban sector also known as the industrial or modem sector. (Bourguignon and Morrisson 1998) This treatment of developing economies in modern development theory can be traced back to the 1954 paper of Sir Arthur Lewis, ‘Economic Development with Unlimited Supplies of Labour’, although the genesis of the term “dual economies” is found in a 1953 paper by Boeke (Kanbur and McIntosh 1989).

Two models of the dual economy have dominated the field. In particular these models have been central in the discussion of labour markets, the interaction between rural and urban labour markets and wages, and by implication the discussion of income distributions. These two models are the Lewis and the Harris-Todaro models. (Bourguignon and Morrisson 1998) The Lewis model depicts the developing economies as being composed of a capitalist/modern sector and a subsistence/traditional sector. The links between the two sectors are as follows (Lewis 1954)

Firstly, the subsistence sector provides an unlimited supply of labour for the capitalist sector, i.e. the capitalist sector can hire whatever amount of labour that is required at the prevailing wage. Then, in the subsistence sector, because production is organised on a family basis, labour appropriates all production and an individual’s share of income is roughly equal to the average productivity of labour.

Consequently, wages in the capitalist sector are anchored by incomes in the subsistence sector and therefore the capitalist has an interest in keeping subsistence sector incomes low. Then, migration from the traditional sector to the modern sector is induced by a wage premium that keeps industrial wages some small constant above traditional sector income.

In addition, Wages in the capitalist sector begin to rise only when one of the following occurs; the supply of labour is exhausted, forcing capitalists to compete for workers; the subsistence sector improves its physical productivity thus increasing the average product of labour in the rural area; or the capitalist sector makes substantial demands on the product of the subsistence sector, hence driving up agricultural prices and the return to each unit of subsistence labour.

Finally, because urban wages are determined by conditions in the subsistence sector, where incomes are constant and low, all gains from an increase in productivity accrue to the capitalist in the form of profits. Because capitalists have a lower propensity to consume than workers, an increasing share of the national product is saved and hence investment increases. The result of this is an expansion of the capitalist sector. (Lewis 1954)

The implication of the above, which is similar to that of the traditional classical model, is that a greater share of national income accrues to the capitalist sector as it expands. It therefore follows that there would be a fall in the rural share of national income as growth occurs. Furthermore, because there is a premium needed to induce migration, there will always be a gap between urban and rural incomes, although presumably a rather small one. (Harris and Todaro 1970)

Factors Determining the Incidence of Rural Poverty

The incidence of poverty among the landless and the peasantry in the countrywide may be reduced either by the expansion of employment in the modern sectors, reducing population and surplus labour in the countryside, and hence the man/land ratio; or by enrichment within the agricultural sector by increasing the resources available to the rural poor, and the yields earned on those resources, expanding agricultural employment opportunities, or improving the terms at which the agricultural sector exchanged commodities.

Initially, emphasis was placed on the first route. Lewis (1954) presented the peasant economy as a subsistence economy, characterized by surplus labour and operating without any capital at all. It served as little more than a holding tank for the reserves of labour which could be made available to the growing capitalist sectors, which alone were capable of investment and growth.

Rural poverty would be reduced by encouraging and stimulating the capitalist sectors to expand. Growth would eventually eliminate the surplus reserves of labour and the incidence of rural poverty would decline as the land available per capita within the peasant sector rose and population fell. Once the surplus had been eliminated further increases in labour demand would be translated into wage increases and a rising standard of living for the employed working class.

From this perspective, Lewis viewed increasing inequality as necessary to reduce the incidence of poverty and underemployment. Lewis writes, “The central fact of economic development is that the distribution of income is altered in favour of the saving class” (Lewis 1954, p.147) and “We are interested not in the-people in general, but only in the 10 percent of them with the largest incomes, who in countries with surplus labour receive up to 40 per cent of the national income. The remaining 90 per cent of the people never manage to save a significant fraction of their incomes” (Lewis 1954, p.146).

Lewis was more concerned with the incidence of poverty than the distribution of income. But he argued that poverty could only be reduced by increasing the profits of the capitalists; only the opportunity to earn higher profits could induce them to change a society capable of saving and investing 4 or 5 per cent of national income into one capable of rates, of 15 to 20 per cent. (Kanbur and McIntosh, 1989).

Lewis Theory Explained

Lewis says because there is surplus labour in the agricultural sector you can therefore get all the labour you want at a fixed wage. If this is the case then as more labour is employed the marginal product of labour goes down. For example if there is one machine, and one man can produce 200 units of output from it, then by adding a second worker you may be able to produce 380 units of output and finally by adding a third worker you can produce 480 units of output. (Lewis 1954) You can see that by the time the third worker is added you are receiving much less than you would expect for a lot more in wage. Therefore firms will only hire up to the point where the marginal product of labour equals the wage.

Figure 1. Production Possibility Frontier for the Urban sector

Assuming capitalists do not consume their profit but rather want to reinvest in order to gain more we can see that from the PPF above, reinvesting shifts the PPF from N1, to N2. Employment in this case raises from M1 to M2 as well. In fact, this will keep on until there is no surplus labour in the economy and eventually the capitalists will have to start paying higher wages in order to get more labour. Now the majority of the people work in the industrial sector and the agricultural sector has now been modernised with wage equal to the marginal product of labour.

Now suppose as workers go to the industrial sector the wages rise. The profits of the capitalist will decrease. Therefore, for this theory to work the wage must be constant. With little profits the capitalist sector will be small with not much to invest. The wage may rise due to trade unions demanding higher wages and government intervention may also cause wages to rise, which would again reduce capitalists’ profits.

One reason endogenous to the model for rising wages would be the following. Clearly, the agricultural level will be getting smaller. However, the demand for food will still be there and this can only be met by the agricultural sector. Therefore the demand for food will increase but the supply will decline causing food prices to go up. So in order to buy food you will need bigger wages. (Ranis and Stewart 1999)

Therefore, as we have already established the wage must remain constant in order for the economy to grow so how can this be achieved? Only by increasing supply in the agricultural sector at the same time. Agricultural reinvestment is a pre-requisite for the industrial revolution. If it doesn’t happen then the price of food will not remain constant.

Figure 2. Correlation between Productivity and Employment in Sectors

A further way to look at the model is to use the above diagram. Here point a is considered to be the equilibrium. P.dxa/dna is the productivity or marginal product of the agricultural sector and dXm/dNm is the productivity of the manufacturing sector. Remember, migration occurs when there is a wage differential.

So at point b we can see P.dxa/dna > dXm/dNm which implies the wage is greater in the agricultural sector than in the manufacturing sector, so people will migrate there until we reach point a. The reverse is also true. When dXm/dNm > P.dxa/dna then the wage is greater in the manufacturing sector and so the rural sector inhabitants will migrate to the urban sector from point c up until they reach point a.

Implications of the Lewis Model:
• Saving and investment are vital for growth. You need capitalists in order to grow!
• Structural transformation of the economy is crucial if you want to grow.
• Income distribution will be uneven to start with. The share of wages will increase in the early stages causing the share of profits to fall. In the long run however, this should even out.
• Agricultural development is a pre-requisite to industrial development. Without agricultural surplus, the urban economy cannot grow.
• Rural to Urban migration is desirable and conclusive to development. Or in other words, urbanisation is a good thing. (Harris and Todaro 1970)

Theoretical and Empirical Evidence

By the mid 1960’s studies were appearing which documented the limited contribution to employment creation modern sector expansion was making. Little et al. (1970) confirmed that in almost all the countries for which adequate data were available, modern sector employment growth was significantly lagging behind output growth, as modem sector expansion was becoming increasingly capital intensive. In many cases modern sector expansion was unable to put a noticeable dent in the reserves of surplus labour in the countryside. Little et al. (1970) and Fields (1984) argued that the pro-industrialisation policies typically involved subsidised credit overvalued exchange rates and capital subsidies, which encouraged the substitution of capital for labour.

Fields (1984) cites the examples of South Korea, Taiwan, Hong Kong, and Costa Rica as countries which did experience a rapid expansion of modem sector employment. In the Asian examples Fields argues that a successful shift from import substitution to export promotion, coupled with market determined wages encouraged the rapid expansion of employment opportunities in the modern sectors. The necessity of ‘getting factor prices right’ is the major conclusion drawn from these studies.

Lipton (1977) extends this thinking into a more general argument as to why the rural poor will remain both rural and poor. The major conflict in LDC’s, Lipton argues, is not between workers and capitalists, but rather between the city and the countryside. Almost inevitably the city wins, captures the government, and thereby ensures that most resources for development are directed towards urban interests. Despite high rates of return on investment in small-scale agriculture, urban, capitalists and workers form an alliance that includes large capitalist farmers to squeeze the peasant economy. Pro-urban, pro-industry development strategies (i.e. Lewis) are merely efforts of the elites to cloak their self-interest in a guise, which appears to promote national development.

Lipton challenges two of Lewis’s major assumptions; one, that yields cannot be significantly increased within the peasant sector and two, that opportunities for viable employment creation within the peasant sector do not exist. Hayami and Ruttan (1985) confirm that once supply constraints facing small farmers have been overcome this sector may indeed become the most significant source of employment creation, and hence bring an economy closer to the elimination of the labour surplus which is the key to overcoming the disappointing distributional consequences of growth. The small farm sector is also seen as capable of significant output expansion as well.

There are probably few LDC’s (and DC’s for that matter), which do not exhibit some degree of ‘urban bias’ with respect to public expenditures on social overhead capital, education, health and other services. Many have also imposed various forms of ‘price twists’; either by cheap food policies or policies which artificially raise the prices of urban based manufactured goods purchased by rural consumers. Lipton’s work is part of a research thrust, which expanded rapidly over the 1970’s which focussed more directly on the problems within the agricultural sector.

A small, open economy is divided into three subsectors; a booming export sector (B). a lagging sector (L) which is comprised of other tradable commodities-and a Non-tradable sector (N). Output is produced by one factor specific to each sector and one mobile factor, labour. Factor prices are flexible. The prices of the first two are given exogenously by world prices and the latter is determined by domestic supply and demand.

The model is real one; monetary considerations are ignored. The exchange rate between tradables and non-tradables may change, but not between B and L, thus B and L may be aggregated into T, the traded goods sector. The basic model described below is taken from Corden and Neary (1982). In such a model, there are two main effects of a boom in B, which we may assume is derived from a costless discovery of new natural resources. L shall be taken to include peasant sector export and import competing goods, and N includes non-traded services and the protected manufacturing sector.

The pre-boom equilibrium corresponds to point A in Figure 3 below. Figure illustrates the labour market. OnOt, measured on the horizontal axis, gives the labour supply. Labour input into N is measured by the horizontal distance from On, and labour input into T is measured by the horizontal distance, from Ot. The wage rate, measured on the vertical axis, is measured in units of output from the lagging sector L, which may be taken to be food. Ll is the labour demand in the lagging sector. Lt is the labour demand schedule for the entire traded goods sector. Ln is the labour demand schedule for the non-traded sector.

The impact on the lagging sector is clearer. The real price of output in the lagging sector has declined, relative to the non-traded sector. Output also falls, both because of the initial resource movement effect and also because of the spending effect. The spending effect raises the demand for non-tradables, and hence the demand for labour in this sector. This is shown by Ln in Figure 3. The subsequent rise in wages causes employment and output in the L sector to decline.

Ln’
Ln
L Lt Ll
W2
W1
A
W0

Figure 2.1. The Labour Market in a Developing Economy

Consider the effect of natural resource discovery. Initially we assumed relative prices unchanged. In Figure 3, Lt increases to Ll. This raises the equilibrium wage rate, and employment in B increases while employment in L and N decline, as a result of the higher wage. Employment in L has declined from D to E.

The impact on the lagging sector is clearer. The real price of output in the lagging sector has declined, relative to the non-traded sector. Output also falls, both because of the initial resource movement effect and also because of the spending effect. The spending effect raises the demand for non-tradables, and hence the demand for labour in this sector. This is shown by Ln in Figure 3. The subsequent rise in wages causes employment and output in the L sector to decline.

If the lagging sector is taken to include much of the peasant sector, then the bauxite boom will contribute both towards reducing the prices of peasant products, in terms of manufactured goods, and falling peasant production. If the booming sector employs marginal amounts of domestic resources, or it’s existence docs not significantly affect the amount of labour available to other sectors, then the resource movement effect disappears, and only the spending effect is left.

Post World-War II studies yield the reality of employment and migration in the most informative way. Industrialisation proved to be neither a solution for employment nor for economic development. Only in Asian developing countries did growth seem to rise with industrialisation. (This is perhaps down to the fact that labour was at a great surplus in these countries and these employees were well educated and skilled. Therefore the human capital and labour could be fully exploited.)

Outside of Asian, however, the expansion of industrial employment was disappointing at best. In developing countries experiencing rapid industrialisation the growth of the industrial employment was generally about half the rate of growth of industrial output causing economists to take a more critical look at the role of industrialisation and the problems of employment and migration. Often this is because many developing countries have an abundance of natural resources but lack the human capital. And so industrialisation is capital and resource intensive rather than labour intensive.

To test this theory I thought I would look at the role of industrialisation and economic growth in a developing country to see how it compares.
Zimbabwe 2000 2001 2002 2003 2004 2005
Agriculture, value added (% of GDP) 18,49 17,23 13,73 15,58 17,8 22,35
GDP growth (annual %) -7,9 -2,7 -4,4 -10,4 -4,2 -7,1
Industry, value added (% of GDP) 24,98 22,24 20,4 19,99 22,71 27,96

You can see even from this rough data that I have collected from the World Banks’ website (http://devdata.worldbank.org/data-query/) that as Zimbabwe’s industrial sector has grown it’s economy has not although I could not find any specific data relating to there unemployment levels over this period.

Finally, empirical surveys of rural labour markets in developing countries generally do not support the assumption of surplus labour. Therefore, empirically neither model seems to be correct if you apply it to the real world although both provide a good base from which to start further theories. What has been specifically found in developing countries is that small peasant farmers tend to participate in rural labour markets both by supplying and demanding labour. And that this supply and demand of labour varies considerably over the different crop seasons. Therefore there are labour shortages as well as labour surpluses depending upon the season.

Conclusion

Modern development theorists have retained the basic structure of a dual economy but they have abandoned the social relations that were so important to their progenitors. Instead of a division between classes we are left with a spatial division in the economy, which may or may not correspond with class divisions.

This neglect of social relations has locked modem development theorists into a duality (rural-urban), which may well exist at some levels but it obscures, other important differences and relationships that are present in less industrialised economies. This movement toward a non-class approach to income distribution is probably attributable to the dominance of supply and demand theories in the discipline. This approach requires that agents in the market be homogeneous. Here labour can hire capital, just as capital hires labour.

References
Bourguignon, F. and C. Morrisson (1998) ‘Inequality and development: the role of dualism’, Journal of Development Economics, 57(2):233-357.
Corden, W.M. and J. Neary, “Booming Sector and De-industrialization in a Small, Open Economy,” Economic Journal vol. 92. 1982.
Fields, G. S. (2004) ‘Dualism in the Labor Market: a Perspective on the Lewis Model. after Half a Century’, The Manchester School, Vol. 72, pp. 724–735.
Fields, G. (1980) Poverty, Inequality, and Development. Cambridge University Press, New York.
Hayami, Y. and Ruttan, V.W. (1985) Agricultural Development. An International Perspective, 2nd.edition. Johns Hopkins Press, Baltimore.
Harris, J. R. and M. P. Todaro (1970) ‘Migration, Unemployment and Development: A Two-Sector Analysis’, American Economic Review, vol. 60, pp. 126-142.
Lewis, W. A. (1954) ‘Economic Development with Unlimited Supplies of Labour’, The Manchester School, vol. 22, no. 2, pp. 139-191
Lipton. M. (1977). Why Poor People Stay Poor: Urban Bias in World Development. Temple Smith, London.
Little, I.M. (1984) Economic Development. Basic Books, New York.
Kanbur, R. and J. McIntosh (1989) “Dual Economies” in Eatwell et al (eds.) New Palgraves, W.W. Norton, New York.
Ranis, G. and F. Stewart (1999) ‘V-goods and the Role of the Urban Informal Sector’, Economic Development and Cultural Change, vol. 47, no. 2, pp. 259-288.
Samuelson, P. (1977) Economics, McGraw-Hill, Inc.
Smith, Adam (1976) An Inquiry Into The Nature and Causes of The Wealth of Nations, Oxford University Press, Oxford.
Weldon J.C. (1983) “The Classical Theory of Distribution” in Asimakopulos, A. (ed.) Theories of Income Distribution, Kluwer Academic Publishers, Boston.