MacroScan (Printable Version) - Employment and Poverty among India's Neighbours

It is notable that despite significant differences across the countries of South Asia in terms of size, resource endowment, particular social and political configurations, all of these economies share certain structural characteristics. These include: the presence of a high degree of underemployment; a strong dualism between organized and unorganized sectors, especially in manufacturing, which sometimes (but not always) translates into the dualism between large-scale and small-scale sectors; the continuing significance of agriculture as a major employer; the emergence of services as the largest employers, often as a refuge sector; the involvement of a large share of the workforce in what is essentially low-productivity employment.

But in addition to these, what is more remarkable is the apparent synchronicity of policies and processes across the region, despite very differing social and political pressures. All the economies of the region, for the first few decades after independence, adopted import-substituting industrialization strategies, with the attendant development of some industry and associated dualism in the economy, as well as regulation of much economic activity.

From the 1980s onwards, all of them moved, in varying degrees, to a strategy of development based on export-orientation, liberalization and privatization based on the marketist neoliberal economic paradigm. The process could be said to have started in South Asia with the Sri Lankan government of Jayawardene moving towards liberalization and dismantling of the earlier universal food security system, in the late 1970s and early 1980s.

Subsequently, and more strongly in the early 1990s, all the governments in the region (barring that of Nepal, which had very a different position) went through fairly comprehensive policies of internal liberalisation-reduction of direct state responsibility for a range of goods and services-and privatization.

By the turn of the century, most of the important economies in South Asia had undergone

substantial reduction in direct state control in terms of administered prices, regulation of economic activity;
privatization of state assets, often in controversial circumstances;
rationalization (usually also a euphemism for reduction) of direct and indirect tax rates that became associated with declining tax-GDP ratios;
attempts (typically unsuccessful) to reduce fiscal deficits which usually involved cutting back on public productive investment as well as certain types of social expenditure, reducing subsidies to farmers and increasing user charges for public services and utilities;
trade liberalization, involving shifts from quantitative restrictions to tariffs and typically sharp reductions in the average rate of tariff protection;
financial liberalization involving reductions in directed credit, freeing of interest rate ceilings and other measures that raised the cost of borrowing, including for the government;
moving to market determined exchange rates and liberalization of current account transactions;
allowing some degree of capital account liberalization, including easing of rules for Foreign Direct Investment, allowing non-residents to hold domestic financial assets and providing easier access to foreign commercial borrowing by domestic firms.

This commonality of policy experience meant in turn that outcomes were also quite similar, despite the very different initial conditions in different economies. Some of these outcomes are discussed in detail below.

Pakistan
The economy of Pakistan     is viewed internationally as having a reasonably good rate of output growth. This perception is more a reflection of the general slowdown in economic growth across the world, than in any improvement in Pakistan's growth performance per se. It is true that for the period 1960–90, Pakistan's growth was high for a low-income country, at around 6 per cent per annum with a 2 per cent variation. However, the 1990s involved a significant deceleration of growth in Pakistan, especially in certain sectors such as manufacturing.

Further, such growth has been associated with very inadequate performance in terms of human development indicators. William Easterly (2001) has argued that Pakistan's pattern is indicative of 'growth without development', because despite its 'respectable' per capita growth over the second half of the twentieth century, the country has 'systematically underperformed' on most social and political indicators, such as education, health, sanitation, fertility, gender equality, corruption, political instability and violence, and democracy. Significantly, such output growth has also been associated with much lower employment growth, at the trend rate of only 2 per cent per annum for the long period of 1960–99.

During the recent years, a number of features of the economic growth process in Pakistan are worth noting. First, is the very high degree of volatility of growth, with very significant fluctuations over the years, and a trend deceleration evident in the 1990s. Second, is that this growth has been based largely on unsustainable public expenditure using a build-up of public debt that has already reached problematic levels. Total debt-servicing (of external and internal debt together) already accounts for more than 71 per cent of current government revenues, and future expansion cannot rely on similar debt-driven public spending alone.

Third, is the very high presence of underemployment or disguised unemployment. The Labour Force Survey data suggest underemployment rates of 13.3 per cent in rural areas and 6.3 per cent in urban areas in the mid-1990s, but these are likely to be underestimates.

Fourth, is the fact that there appears to be relatively less direct relation between growth and employment generation. As Chart 1 indicates, output growth was relatively low in the 1970s. It increased in the 1980s and dropped again in the 1990s. But employment growth followed the opposite pattern, being at its highest at 3 per cent over the 1970s and dropping to 2 per cent in the next two decades. Nomaan Majid (2002) finds a total break between growth and employment in the period after the mid-1980s, and particularly in the period after the imposition of an International Monetary Programme (IMF)-induced Structural Adjustment Programme in 1987–88.



It is apparent that in terms of output growth, manufacturing (Chart 3) was the lead sector over all three periods, and agriculture (Chart 2) contributed progressively less over time. In terms of employment growth, manufacturing led during the low growth phase of the 1970s, and in the 1990s when output growth rates declined once again, employment growth in manufacturing was actually negative.





Majid (2000) suggests that the overall break between output and employment therefore seems to be associated more with the manufacturing sector, and that agriculture and construction (Chart 4) may have become residual 'refuge' sectors in the most recent period. This would also explain negative productivity growth in construction in the 1990s.



This analysis implies that the manufacturing sector is the key to the explanation of the poor employment performance of the past. Manufacturing in Pakistan, as in most other developing nations, is characterized by a high degree of dualism. There is a large-scale sector that dominates output (producing two-thirds of the value added in manufacturing) but employs only 17 per cent of manufacturing workers, and a small-scale sector that dominates employment (with 83 per cent of the manufacturing workforce) but accounts for only one-third of the manufacturing value added.

The output and employment shares of these two categories have been remarkably stable over time. The small-scale sector operating under major and increasing constraints and with huge disadvantages, has been relatively moribund in the last decade, and shows all the characteristics of a refuge labour sector. Meanwhile, the large-scale sector has been plagued by excess capacity (due to deficient aggregate demand resulting from deflationary structural adjustment policies, and import penetration) as well as by increasing capital intensity and in capital productivity due to newer technologies that have had the effect of reducing labour demand.

So, much as had occurred in India over the same period, investment and output growth in manufacturing in Pakistan tended to be capital-augmenting and labour-displacing. Since manufacturing was the lead sector in employment generation, this then affected the employment possibilities elsewhere in the economy, and explains both the persistence of low-productivity employment in the residual sectors and the low and declining rates of labour force participation in Pakistan.

The significance of deflationary macro-policies in affecting both growth and employment in Pakistan has already been mentioned. A very major and direct role was played in this case by the constraints imposed on public investment. The investment–GDP ratio declined from 17.3 per cent in 1998–89 to 14.7 per cent in 2000–01, and this was entirely due to the collapse in public investment from 8.5 per cent of GDP to 5.6 per cent over the same period.

Private investment, which is strongly interlinked with public investment and expenditure, faced a deficiency of demand as a result, and did not rise to meet the emerging slack. In addition, various other elements of the structural adjustment programme operated to reduce average growth rates, accelerate inflation, and thereby increase unemployment and poverty. The standard package of structural reforms included privatization of public assets, ceilings on wages and employment in the public sector, cuts in subsidies, cuts in development expenditure, including on 'social sectors', increases in user charges for public utilities and services and frequent devaluation.

This last feature also had the unintended consequence of reducing the inflow of remittances from foreign workers, which has been an important source of sustenance of Pakistan's balance of payments. (Anwar 2001). Thus, ironically, the macroeconomic strategy based on Structural Adjustment Programmes imposed and approved by the IMF and World Bank supposedly to change the structure of the economy so as to improve the balance of payments, control inflation and revive growth, had the opposite effects in practice.

Inadequate employment generation and persistence of low productivity employment in most sectors inevitably feed into and even increase levels of poverty. In Pakistan, there is general agreement that the incidence of poverty has increased over the 1990s, as the combination of deflationary macro-economic measures and de-industrialization following upon trade liberalization has made itself felt. Chart 5 describes the broad trends in poverty measured by the head count ratio.



It is noteworthy that the Structural Adjustment Programme inspired by the IMF (which the IMF itself termed as a 'success')    began in 1987–88, after which the incidence of poverty has been increasing. There are various reasons behind this, most of which relate to the direct and indirect effects of the macro-economic strategy as elaborated below. What is relevant here is that this intertwining of low employment generation and increasing poverty has been an important feature of the experience of the last decade and more so in Pakistan.

Bangladesh
Much like other countries in the region, Bangladesh also adopted the standard range of neo-liberal economic reforms beginning in the 1980s and much more comprehensively in the 1990s. These included the usual combination of expenditure-reducing and expenditure-switching policies, along with institutional and legislative changes aimed at ensuring greater freedom of market functioning.

The results of this strategy appear to have been mixed. There was clearly greater macroeconomic stability in terms of reduced rates of inflation (from an average of 9.9. per cent per annum in the early 1980s to an average of 5.6 per cent per annum by the end of the 1990s). However, the trend rate of growth of GDP in constant prices appears to have been remarkably stable, at around 4 per cent, suggesting that the reforms package had very little effect on the trend rate of growth.

Also, trends in the fiscal deficit (as share of GDP) appear to have been substantially post reform, in the 1990s, compared to the 1980s. This also reflects the decline in tax revenues consequent upon fiscal reforms and trade liberalization.

Muqtada (2003, page 8) has concluded that 'stabilization and the SAPs, to the extent implemented, do not seem to have produced any perceptible growth dividends'. It is notable that while the supposed 'fundamentals' of the economy have improved, these have failed to increase rates of investment and growth. However, growth rates per capita do show significant increase, mainly because of the demographic transition involving declines in rates of population growth (from 2.1 per cent in 1990 to only 1.6 per cent in 2000).

The most recent Labour Force Survey suggests that employment growth grew at around 3 per cent per annum in the period 1995–2000. While this is still below the rate of increase of the labour force (estimated to be 3.4 per cent by the same survey) it is higher than in the earlier period.

However, there are doubts about how and in which sectors such a jump in employment occurred. While the new export sector of readymade garments has provided an important source of new employment (especially for women) total employment in aggregate manufacturing has actually declined, in both relative and absolute terms.

There has been significant de-industrialization, particularly in the traditional sectors that have suffered from import penetration. Open unemployment has risen from 1.8 per cent of the labour force to 4.9 per cent, while underemployment in 2000 was estimated to be very high at around 31 per cent.

What is more remarkable in the case of Bangladesh is the fact that changes in the structure of output of the economy have been completely different from changes in the sectoral composition of employment. These are shown in Chart 6 and Chart 7.





It is apparent from these charts that while the GDP share of agriculture fell by nearly half, its share of employment actually increased, contrary to the standard expectation with respect to the process of development. Indeed, the distribution of employment reflects hardly any structural change in the economy, while the composition of GDP indicates a more usual process of declining agriculture and increasing industry, albeit with a very large role played by services.

Average productivity in agriculture remains low and wages in that sector have been lower than in manufacturing, with the gap widening. Therefore agriculture is clearly a residual sector in the context of Bangladesh, absorbing labour that cannot obtain productive employment elsewhere. It is possible that services have been playing a similar role.

The poor employment generation in the manufacturing sector reflects a combination of declines in the traditional manufacturing sectors and (given the rising share of aggregate output) increases in labour productivity which are the result of sectoral shifts and changing technology. The latter tends to be confirmed by the rising real wages in manufacturing. This conforms to the pattern we have already observed in other countries of South Asia, where the competitive pressures resulting from greater openness have involved technological changes which are on the whole labour-displacing.

Poverty is estimated to have declined in the latter part of the 1990s, although the rural incidence remains high at around 30 per cent of the rural population.

There is a common perception that the micro-credit delivery systems in Bangladesh has operated to provide a cushion for poor households in case of shocks such as crop failures, floods and other natural disasters. It has also helped to improve the relative position of women. However, the features of micro-credit (short-term, relatively small amounts, groups lending pressure for prompt repayment) mean that it has not contributed much to asset creation among the poor or to sustained employment generation. (Centre for Policy Dialogue 2001)

The expansion of public transport infrastructure, especially road- networks in the 1980s, may have contributed to subsequent rural development which in turn assisted the reduction of poverty in that later period. However, trade liberalization had the counter effect of reducing the viability of many small producers, so the net effect of all the policy changes over the period is not clear.

It is likely that some of the effects of openness were adverse for livelihood and therefore it led to poverty, and these were to some extent mitigated by the spread of public transport networks and the availability of micro-credit.

Sri Lanka
At first glance, Sri Lanka apparently shows reasonably good economic growth and employment generation performance from the early 1990s. This is somewhat surprising, since the country was among the least stable due to the intensification of internal conflict, and has also undergone similar policies of neoliberal adjustment and privatization from the 1980s onwards.

The country had to suffer major violent conflicts over the last decade, amounting at times to a civil war, especially in the northern Tamil-dominated region. This not only meant the destruction of many lives and of physical and social infrastructure, but also created uncertain conditions for private investment and involved massive increases in military spending by the government.

In fact the growth of the Sri Lankan economy-at an average of more than 5 per cent per annum over the 1990s-exhibited substantial volatility and there are also doubts about its sustainability, because of the large build-up of public debt it has been associated with. By the turn of the decade the fragile nature of the growth process and the increased external vulnerability of the Sri Lankan economy were exposed, when there was a mini balance-of-payments crisis, resulting from a temporary decline in aid inflows and export receipts. In 2001, GDP fell by 1.3 per cent.

Sri Lanka underwent neo-liberal economic reform packages similar to those of other countries in South Asia, but from an even earlier period, starting in the late 1970s. The process was marked in the 1980s with the replacement of a universal system of food distribution with targeted food stamps and other changes deigned to reduce the welfare redistributive schemes of the state, and continued with an intensified programme of fiscal retrenchment and privatization of state assets, including plantations, from the early 1990s.

The economic policies adopted during the last two decades were oriented towards accelerating 'growth' through liberalization, export orientation and privatization, with the assumption that growth would trickle down and reduce poverty. However, even the World Bank, which was actively associated with promoting these policies, now admits that neither adequate growth nor poverty reduction have been achieved during this period (World Bank 2002: 10).

While the rate of growth of output increased in the 1990s, it is noteworthy that it was associated with lower rates of employment expansion than the previous two decades that were characterized by lower rates of aggregate GDP growth (Chart 8). This is precisely the same pattern we have already observed in Pakistan, and to a lesser extent in India.



In terms of structural change, the Sri Lankan economy experienced what would be considered as the 'standard' process of relative decline in the share of the primary sector. However, the share of the secondary sector was broadly stable over the last two decades after declining in the 1970s, and services now account for more than half of GDP . (Chart 9)



In fact, there is more to this structural change than is immediately apparent, since it is the result of a significantly changing profile in agriculture, with a steady decline in the traditional export crops. The share of such crops (tea, rubber and coconut) declined from 34 per cent of value added in agriculture in 1980 to only 20 per cent in 1999. Even rice, the mainstay of traditional agriculture, declined from 27 per cent of value added in agriculture to 18 per cent over the same period, as a range of 'other crops' emerged to dominate. The decline in the plantation crops was reflected in the declining share of the processing of tea, rubber and coconut in total manufacturing, from 34 per cent to 15 per cent, and a corresponding rise in the share of factory industry.

This was also reflected in the changing composition of exports. The share of agricultural exports declined from 62 per cent in 1980 to 21 per cent in 1999, while industrial exports increased from 33 per cent to 77 per cent.

However, these massive shifts in production structure are not reflected in employment patterns: while the share of agriculture in total employment declined from 42 per cent to 36 per cent, the share of manufacturing employment remained absolutely stagnant at only 15 per cent, between 1990 and 1999. As in other South Asian countries, the increase has been in services employment, which is likely to have been operating as the residual or refuge sector once again.

It is true that open unemployment rates have declined substantially. But this is likely to reflect more of the discouraged-worker effect than an actual increase in job availability, given the deceleration in employment expansion that was already noted.

Further, as the Government of Sri Lanka has also noted (2002: 4) 'Among the so-called "employed" are people that have worked for as little as one hour per week in paid employment. This definition hides many who are significantly under-employed and who are in need of full-time productive jobs. (It has been estimated that the number of "under-employed" amount to as many as 20 per cent of the total workforce, or approximately 1.3 million people.) Also included as "employed" are approximately 673,000 people that are classified as "unpaid family workers", many of who would no doubt welcome full-time, paid work if it were available."

The official estimates for Sri Lanka suggest fluctuations and no downward trend in the incidence of poverty by the head count ratio (Chart 10). Further, there is substantial difference across the regions: if the more prosperous western region was excluded, aggregate poverty figures would be much higher, since it is in the range of 41–55 per cent for all the other provinces, according to the higher poverty line. Poverty incidence is highest in the estate sector, at 45 per cent, and lowest in urban Sri Lanka at 25 per cent but in all of these categories, it has stubbornly persisted at its earlier levels despite the more-than-5 per cent growth of the economy.



There is little doubt that this persistence of poverty is related not only to various measures taken by the government which have reduced access to basic needs and public goods over the period from the mid-1980s, but also because the process of liberalization and privatization thus far has simply not generated enough productive employment.

Conclusion
Given the patterns observed above, it is not surprising that increasing employment generation is now the explicit concern in recent planning and policy documents in all of these countries, even India. It is strange however, that while the explicit goal has changed from growth in itself to employment generation, the strategies that are supposed to achieve this essentially involve further doses of neo-liberal marketist reform, rather than policies that would directly affect employment.

Thus, most of the policy statements refer to further privatization, further deregulation of domestic economic activity, further financial liberalization and external capital account liberalization, and further restrictions on fiscal policies. These are precisely the set of policies that, as observed already, have been associated with deceleration of employment in the last decade.

If employment generation is to be the focus of the new policy thrust in the region, then it would actually require a rethinking of these policies, towards more active state intervention in terms of supporting employment-intensive activities through a range of trade, fiscal and financial measures. Without such active involvement, aggregate employment in the region is likely to continue to stagnate, and may even deteriorate with further doses of neoliberal reform.



Themes > Features	20.06.2003


Employment and Poverty among India's Neighbours

It is notable that despite significant differences across the countries of South Asia in terms of size, resource endowment, particular social and political configurations, all of these economies share certain structural characteristics. These include: the presence of a high degree of underemployment; a strong dualism between organized and unorganized sectors, especially in manufacturing, which sometimes (but not always) translates into the dualism between large-scale and small-scale sectors; the continuing significance of agriculture as a major employer; the emergence of services as the largest employers, often as a refuge sector; the involvement of a large share of the workforce in what is essentially low-productivity employment. But in addition to these, what is more remarkable is the apparent synchronicity of policies and processes across the region, despite very differing social and political pressures. All the economies of the region, for the first few decades after independence, adopted import-substituting industrialization strategies, with the attendant development of some industry and associated dualism in the economy, as well as regulation of much economic activity. From the 1980s onwards, all of them moved, in varying degrees, to a strategy of development based on export-orientation, liberalization and privatization based on the marketist neoliberal economic paradigm. The process could be said to have started in South Asia with the Sri Lankan government of Jayawardene moving towards liberalization and dismantling of the earlier universal food security system, in the late 1970s and early 1980s. Subsequently, and more strongly in the early 1990s, all the governments in the region (barring that of Nepal, which had very a different position) went through fairly comprehensive policies of internal liberalisation-reduction of direct state responsibility for a range of goods and services-and privatization. By the turn of the century, most of the important economies in South Asia had undergone substantial reduction in direct state control in terms of administered prices, regulation of economic activity; privatization of state assets, often in controversial circumstances; rationalization (usually also a euphemism for reduction) of direct and indirect tax rates that became associated with declining tax-GDP ratios; attempts (typically unsuccessful) to reduce fiscal deficits which usually involved cutting back on public productive investment as well as certain types of social expenditure, reducing subsidies to farmers and increasing user charges for public services and utilities; trade liberalization, involving shifts from quantitative restrictions to tariffs and typically sharp reductions in the average rate of tariff protection; financial liberalization involving reductions in directed credit, freeing of interest rate ceilings and other measures that raised the cost of borrowing, including for the government; moving to market determined exchange rates and liberalization of current account transactions; allowing some degree of capital account liberalization, including easing of rules for Foreign Direct Investment, allowing non-residents to hold domestic financial assets and providing easier access to foreign commercial borrowing by domestic firms. This commonality of policy experience meant in turn that outcomes were also quite similar, despite the very different initial conditions in different economies. Some of these outcomes are discussed in detail below. Pakistan The economy of Pakistan is viewed internationally as having a reasonably good rate of output growth. This perception is more a reflection of the general slowdown in economic growth across the world, than in any improvement in Pakistan's growth performance per se. It is true that for the period 1960–90, Pakistan's growth was high for a low-income country, at around 6 per cent per annum with a 2 per cent variation. However, the 1990s involved a significant deceleration of growth in Pakistan, especially in certain sectors such as manufacturing. Further, such growth has been associated with very inadequate performance in terms of human development indicators. William Easterly (2001) has argued that Pakistan's pattern is indicative of 'growth without development', because despite its 'respectable' per capita growth over the second half of the twentieth century, the country has 'systematically underperformed' on most social and political indicators, such as education, health, sanitation, fertility, gender equality, corruption, political instability and violence, and democracy. Significantly, such output growth has also been associated with much lower employment growth, at the trend rate of only 2 per cent per annum for the long period of 1960–99. During the recent years, a number of features of the economic growth process in Pakistan are worth noting. First, is the very high degree of volatility of growth, with very significant fluctuations over the years, and a trend deceleration evident in the 1990s. Second, is that this growth has been based largely on unsustainable public expenditure using a build-up of public debt that has already reached problematic levels. Total debt-servicing (of external and internal debt together) already accounts for more than 71 per cent of current government revenues, and future expansion cannot rely on similar debt-driven public spending alone. Third, is the very high presence of underemployment or disguised unemployment. The Labour Force Survey data suggest underemployment rates of 13.3 per cent in rural areas and 6.3 per cent in urban areas in the mid-1990s, but these are likely to be underestimates. Fourth, is the fact that there appears to be relatively less direct relation between growth and employment generation. As Chart 1 indicates, output growth was relatively low in the 1970s. It increased in the 1980s and dropped again in the 1990s. But employment growth followed the opposite pattern, being at its highest at 3 per cent over the 1970s and dropping to 2 per cent in the next two decades. Nomaan Majid (2002) finds a total break between growth and employment in the period after the mid-1980s, and particularly in the period after the imposition of an International Monetary Programme (IMF)-induced Structural Adjustment Programme in 1987–88. It is apparent that in terms of output growth, manufacturing (Chart 3) was the lead sector over all three periods, and agriculture (Chart 2) contributed progressively less over time. In terms of employment growth, manufacturing led during the low growth phase of the 1970s, and in the 1990s when output growth rates declined once again, employment growth in manufacturing was actually negative. Majid (2000) suggests that the overall break between output and employment therefore seems to be associated more with the manufacturing sector, and that agriculture and construction (Chart 4) may have become residual 'refuge' sectors in the most recent period. This would also explain negative productivity growth in construction in the 1990s. This analysis implies that the manufacturing sector is the key to the explanation of the poor employment performance of the past. Manufacturing in Pakistan, as in most other developing nations, is characterized by a high degree of dualism. There is a large-scale sector that dominates output (producing two-thirds of the value added in manufacturing) but employs only 17 per cent of manufacturing workers, and a small-scale sector that dominates employment (with 83 per cent of the manufacturing workforce) but accounts for only one-third of the manufacturing value added. The output and employment shares of these two categories have been remarkably stable over time. The small-scale sector operating under major and increasing constraints and with huge disadvantages, has been relatively moribund in the last decade, and shows all the characteristics of a refuge labour sector. Meanwhile, the large-scale sector has been plagued by excess capacity (due to deficient aggregate demand resulting from deflationary structural adjustment policies, and import penetration) as well as by increasing capital intensity and in capital productivity due to newer technologies that have had the effect of reducing labour demand. So, much as had occurred in India over the same period, investment and output growth in manufacturing in Pakistan tended to be capital-augmenting and labour-displacing. Since manufacturing was the lead sector in employment generation, this then affected the employment possibilities elsewhere in the economy, and explains both the persistence of low-productivity employment in the residual sectors and the low and declining rates of labour force participation in Pakistan. The significance of deflationary macro-policies in affecting both growth and employment in Pakistan has already been mentioned. A very major and direct role was played in this case by the constraints imposed on public investment. The investment–GDP ratio declined from 17.3 per cent in 1998–89 to 14.7 per cent in 2000–01, and this was entirely due to the collapse in public investment from 8.5 per cent of GDP to 5.6 per cent over the same period. Private investment, which is strongly interlinked with public investment and expenditure, faced a deficiency of demand as a result, and did not rise to meet the emerging slack. In addition, various other elements of the structural adjustment programme operated to reduce average growth rates, accelerate inflation, and thereby increase unemployment and poverty. The standard package of structural reforms included privatization of public assets, ceilings on wages and employment in the public sector, cuts in subsidies, cuts in development expenditure, including on 'social sectors', increases in user charges for public utilities and services and frequent devaluation. This last feature also had the unintended consequence of reducing the inflow of remittances from foreign workers, which has been an important source of sustenance of Pakistan's balance of payments. (Anwar 2001). Thus, ironically, the macroeconomic strategy based on Structural Adjustment Programmes imposed and approved by the IMF and World Bank supposedly to change the structure of the economy so as to improve the balance of payments, control inflation and revive growth, had the opposite effects in practice. Inadequate employment generation and persistence of low productivity employment in most sectors inevitably feed into and even increase levels of poverty. In Pakistan, there is general agreement that the incidence of poverty has increased over the 1990s, as the combination of deflationary macro-economic measures and de-industrialization following upon trade liberalization has made itself felt. Chart 5 describes the broad trends in poverty measured by the head count ratio. It is noteworthy that the Structural Adjustment Programme inspired by the IMF (which the IMF itself termed as a 'success') began in 1987–88, after which the incidence of poverty has been increasing. There are various reasons behind this, most of which relate to the direct and indirect effects of the macro-economic strategy as elaborated below. What is relevant here is that this intertwining of low employment generation and increasing poverty has been an important feature of the experience of the last decade and more so in Pakistan. Bangladesh Much like other countries in the region, Bangladesh also adopted the standard range of neo-liberal economic reforms beginning in the 1980s and much more comprehensively in the 1990s. These included the usual combination of expenditure-reducing and expenditure-switching policies, along with institutional and legislative changes aimed at ensuring greater freedom of market functioning. The results of this strategy appear to have been mixed. There was clearly greater macroeconomic stability in terms of reduced rates of inflation (from an average of 9.9. per cent per annum in the early 1980s to an average of 5.6 per cent per annum by the end of the 1990s). However, the trend rate of growth of GDP in constant prices appears to have been remarkably stable, at around 4 per cent, suggesting that the reforms package had very little effect on the trend rate of growth. Also, trends in the fiscal deficit (as share of GDP) appear to have been substantially post reform, in the 1990s, compared to the 1980s. This also reflects the decline in tax revenues consequent upon fiscal reforms and trade liberalization. Muqtada (2003, page 8) has concluded that 'stabilization and the SAPs, to the extent implemented, do not seem to have produced any perceptible growth dividends'. It is notable that while the supposed 'fundamentals' of the economy have improved, these have failed to increase rates of investment and growth. However, growth rates per capita do show significant increase, mainly because of the demographic transition involving declines in rates of population growth (from 2.1 per cent in 1990 to only 1.6 per cent in 2000). The most recent Labour Force Survey suggests that employment growth grew at around 3 per cent per annum in the period 1995–2000. While this is still below the rate of increase of the labour force (estimated to be 3.4 per cent by the same survey) it is higher than in the earlier period. However, there are doubts about how and in which sectors such a jump in employment occurred. While the new export sector of readymade garments has provided an important source of new employment (especially for women) total employment in aggregate manufacturing has actually declined, in both relative and absolute terms. There has been significant de-industrialization, particularly in the traditional sectors that have suffered from import penetration. Open unemployment has risen from 1.8 per cent of the labour force to 4.9 per cent, while underemployment in 2000 was estimated to be very high at around 31 per cent. What is more remarkable in the case of Bangladesh is the fact that changes in the structure of output of the economy have been completely different from changes in the sectoral composition of employment. These are shown in Chart 6 and Chart 7. It is apparent from these charts that while the GDP share of agriculture fell by nearly half, its share of employment actually increased, contrary to the standard expectation with respect to the process of development. Indeed, the distribution of employment reflects hardly any structural change in the economy, while the composition of GDP indicates a more usual process of declining agriculture and increasing industry, albeit with a very large role played by services. Average productivity in agriculture remains low and wages in that sector have been lower than in manufacturing, with the gap widening. Therefore agriculture is clearly a residual sector in the context of Bangladesh, absorbing labour that cannot obtain productive employment elsewhere. It is possible that services have been playing a similar role. The poor employment generation in the manufacturing sector reflects a combination of declines in the traditional manufacturing sectors and (given the rising share of aggregate output) increases in labour productivity which are the result of sectoral shifts and changing technology. The latter tends to be confirmed by the rising real wages in manufacturing. This conforms to the pattern we have already observed in other countries of South Asia, where the competitive pressures resulting from greater openness have involved technological changes which are on the whole labour-displacing. Poverty is estimated to have declined in the latter part of the 1990s, although the rural incidence remains high at around 30 per cent of the rural population. There is a common perception that the micro-credit delivery systems in Bangladesh has operated to provide a cushion for poor households in case of shocks such as crop failures, floods and other natural disasters. It has also helped to improve the relative position of women. However, the features of micro-credit (short-term, relatively small amounts, groups lending pressure for prompt repayment) mean that it has not contributed much to asset creation among the poor or to sustained employment generation. (Centre for Policy Dialogue 2001) The expansion of public transport infrastructure, especially road- networks in the 1980s, may have contributed to subsequent rural development which in turn assisted the reduction of poverty in that later period. However, trade liberalization had the counter effect of reducing the viability of many small producers, so the net effect of all the policy changes over the period is not clear. It is likely that some of the effects of openness were adverse for livelihood and therefore it led to poverty, and these were to some extent mitigated by the spread of public transport networks and the availability of micro-credit. Sri Lanka At first glance, Sri Lanka apparently shows reasonably good economic growth and employment generation performance from the early 1990s. This is somewhat surprising, since the country was among the least stable due to the intensification of internal conflict, and has also undergone similar policies of neoliberal adjustment and privatization from the 1980s onwards. The country had to suffer major violent conflicts over the last decade, amounting at times to a civil war, especially in the northern Tamil-dominated region. This not only meant the destruction of many lives and of physical and social infrastructure, but also created uncertain conditions for private investment and involved massive increases in military spending by the government. In fact the growth of the Sri Lankan economy-at an average of more than 5 per cent per annum over the 1990s-exhibited substantial volatility and there are also doubts about its sustainability, because of the large build-up of public debt it has been associated with. By the turn of the decade the fragile nature of the growth process and the increased external vulnerability of the Sri Lankan economy were exposed, when there was a mini balance-of-payments crisis, resulting from a temporary decline in aid inflows and export receipts. In 2001, GDP fell by 1.3 per cent. Sri Lanka underwent neo-liberal economic reform packages similar to those of other countries in South Asia, but from an even earlier period, starting in the late 1970s. The process was marked in the 1980s with the replacement of a universal system of food distribution with targeted food stamps and other changes deigned to reduce the welfare redistributive schemes of the state, and continued with an intensified programme of fiscal retrenchment and privatization of state assets, including plantations, from the early 1990s. The economic policies adopted during the last two decades were oriented towards accelerating 'growth' through liberalization, export orientation and privatization, with the assumption that growth would trickle down and reduce poverty. However, even the World Bank, which was actively associated with promoting these policies, now admits that neither adequate growth nor poverty reduction have been achieved during this period (World Bank 2002: 10). While the rate of growth of output increased in the 1990s, it is noteworthy that it was associated with lower rates of employment expansion than the previous two decades that were characterized by lower rates of aggregate GDP growth (Chart 8). This is precisely the same pattern we have already observed in Pakistan, and to a lesser extent in India. In terms of structural change, the Sri Lankan economy experienced what would be considered as the 'standard' process of relative decline in the share of the primary sector. However, the share of the secondary sector was broadly stable over the last two decades after declining in the 1970s, and services now account for more than half of GDP . (Chart 9) In fact, there is more to this structural change than is immediately apparent, since it is the result of a significantly changing profile in agriculture, with a steady decline in the traditional export crops. The share of such crops (tea, rubber and coconut) declined from 34 per cent of value added in agriculture in 1980 to only 20 per cent in 1999. Even rice, the mainstay of traditional agriculture, declined from 27 per cent of value added in agriculture to 18 per cent over the same period, as a range of 'other crops' emerged to dominate. The decline in the plantation crops was reflected in the declining share of the processing of tea, rubber and coconut in total manufacturing, from 34 per cent to 15 per cent, and a corresponding rise in the share of factory industry. This was also reflected in the changing composition of exports. The share of agricultural exports declined from 62 per cent in 1980 to 21 per cent in 1999, while industrial exports increased from 33 per cent to 77 per cent. However, these massive shifts in production structure are not reflected in employment patterns: while the share of agriculture in total employment declined from 42 per cent to 36 per cent, the share of manufacturing employment remained absolutely stagnant at only 15 per cent, between 1990 and 1999. As in other South Asian countries, the increase has been in services employment, which is likely to have been operating as the residual or refuge sector once again. It is true that open unemployment rates have declined substantially. But this is likely to reflect more of the discouraged-worker effect than an actual increase in job availability, given the deceleration in employment expansion that was already noted. Further, as the Government of Sri Lanka has also noted (2002: 4) 'Among the so-called "employed" are people that have worked for as little as one hour per week in paid employment. This definition hides many who are significantly under-employed and who are in need of full-time productive jobs. (It has been estimated that the number of "under-employed" amount to as many as 20 per cent of the total workforce, or approximately 1.3 million people.) Also included as "employed" are approximately 673,000 people that are classified as "unpaid family workers", many of who would no doubt welcome full-time, paid work if it were available." The official estimates for Sri Lanka suggest fluctuations and no downward trend in the incidence of poverty by the head count ratio (Chart 10). Further, there is substantial difference across the regions: if the more prosperous western region was excluded, aggregate poverty figures would be much higher, since it is in the range of 41–55 per cent for all the other provinces, according to the higher poverty line. Poverty incidence is highest in the estate sector, at 45 per cent, and lowest in urban Sri Lanka at 25 per cent but in all of these categories, it has stubbornly persisted at its earlier levels despite the more-than-5 per cent growth of the economy. There is little doubt that this persistence of poverty is related not only to various measures taken by the government which have reduced access to basic needs and public goods over the period from the mid-1980s, but also because the process of liberalization and privatization thus far has simply not generated enough productive employment. Conclusion Given the patterns observed above, it is not surprising that increasing employment generation is now the explicit concern in recent planning and policy documents in all of these countries, even India. It is strange however, that while the explicit goal has changed from growth in itself to employment generation, the strategies that are supposed to achieve this essentially involve further doses of neo-liberal marketist reform, rather than policies that would directly affect employment. Thus, most of the policy statements refer to further privatization, further deregulation of domestic economic activity, further financial liberalization and external capital account liberalization, and further restrictions on fiscal policies. These are precisely the set of policies that, as observed already, have been associated with deceleration of employment in the last decade. If employment generation is to be the focus of the new policy thrust in the region, then it would actually require a rethinking of these policies, towards more active state intervention in terms of supporting employment-intensive activities through a range of trade, fiscal and financial measures. Without such active involvement, aggregate employment in the region is likely to continue to stagnate, and may even deteriorate with further doses of neoliberal reform.



© MACROSCAN 2003