COLUMN

The liberalisation fallout

The experience of Bangladesh in the 1990s points to some of the pitfalls of indiscriminate liberalisation and holds out lessons for other countries including India.

JAYATI GHOSH

ONE of the curious byproducts of the East Asian financial crises - as well as the subsequent eruptions in Russia and Brazil - is the widespread perception that, while capital account liberalisation may be problematic for developing countries, trade liberalisation is always beneficial. The arguments in favour of this view are that such liberalisation forces domestic production to become more competitive and efficient, and provides domestic consumers with a wider range of cheaper products. But in reality there can be very different results, especially if the result is domestic de-industrialisation.

The experience of Bangladesh in the 1990s points to some of the pitfalls of indiscriminate liberalisation. The country experienced very substantial economic liberalisation over this period, with sweeping deregulation of industry accompanying large-scale trade liberalisation and reduction and/or elimination of subsidies, especially for agriculture.

GILLES SAUSSIER/ GAMMA
A scene from rural Bangladesh. Over the last decade and a half, the rural areas have seen a large injection of resources, channelled through government and non-governmental organisations, but there has been a stagnation of productive employment opportunities, despite a slightly increased rate of growth.

Most of these policy changes were donor-driven, which is perhaps not very surprising in an economy in which external (aid-based) financing has typically accounted for around two-thirds of the fiscal deficit and has always financed more than half of the fiscal deficit.

Until recently, the World Bank and the International Monetary Fund had claimed that as a result of liberalisation, the economy now exhibits a new dynamism, supposedly exemplified in higher rates of growth of output, investment and exports.

However, it is now admitted that the growth has been less than was earlier believed, and that the incidence of absolute poverty has remained largely unaffected.

Thus, the average rate of growth of GDP (Gross Domestic Product) at constant prices in the 1990s was 4.9 per cent, which is only slightly higher than in the previous decade. The peak was in 1996-97 at 5.9 per cent, and subsequently the figure has been lower. Agricultural growth decelerated until 1994-95 and recovered thereafter. Industry grew at an average of around 8 per cent, similar to the average of the earlier decade.

Similarly, while both savings and investment rates have improved over the decade, the increase has not been as striking as earlier projected. Gross domestic investment as a share of GDP went up from an average of 12.6 per cent in 1990-93 to 16.8 per cent in 1994-97.

Even this increase was dominantly financed by external resources. Indeed, the savings rate did increase by two percentage points on average, from a mean of 5.6 per cent to 7.6 per cent, but this was still well below the requirement. This underlines the continued dependence of the Bangladesh economy on external financing for investment, which remains a major limitation not only on the potential for economic expansion but also on the policy autonomy of the Government.

The relatively limited dynamism is also evident from the structure of national income and employment. Between 1991-92 and 1997-98, the share of agriculture in GDP decreased by 5 percentage points. But industry's share went up only marginally, from 10.1 per cent to 11.3 per cent. The now-typical developing country pattern, of the share of services increasing first in the face of inadequate industrialisation, was evident in Bangladesh as well. Similarly, the share of manufacturing in total employment actually fell from 11.8 per cent in 1990-91 to a mere 7.5 per cent in 1995-96.

Indeed, throughout the 1990s, employment elasticities of manufacturing output have been falling. Since these have not been compensated for by greatly accelerated output growth, this has meant overall decline in employment in this sector relative to the size of the labour force.

The most dynamic manufacturing sector - that of readymade garments which is dominantly export-oriented - simply cannot absorb the labour that is being shed from these sectors, not to mention the huge reservoir of already underemployed labour that is finding some low-productivity subsistence wage in primary and tertiary sector activities. This is true even at the present very high rates of growth (in excess of 25 per cent) which the readymade garments sector is experiencing, and since even such high growth is unlikely to be sustained for very long, the lack of productive income opportunities in the economy as a whole remains the most pressing problem.

This may be one of the most critical factors explaining the persistence of widespread poverty in the country. The proportion of people living below the poverty line has declined slightly in the first half of the 1990s, from 53 per cent to 51 per cent, but this is still much higher than the 45.9 per cent estimated for 1985-86. Further, the distribution among the those below the poverty line seems to have worsened, making the poorest of the poor even worse off. There has been an improvement in urban poverty. But inequality in consumption expenditure has increased substantially in both rural and urban areas. So in the 1990s in Bangladesh, such growth as has occurred has been more unequally distributed, so that poverty levels and employment generation have remained largely unchanged.

This is a paradox. Over the last decade and a half or more, the rural areas of Bangladesh have seen a large injection of resources, channelled through the Government as well as through non-governmental organisations (NGOs). Government spending on rural areas has gone up substantially since the early 1980s, from around 30 per cent then to as much as 50 per cent by 1994-95. Within this, more was spent on things like rural roads and electricity, as well as primary education. Further, large NGOs such as BRAC, Grameen and Proshikkha and others massively increased their lending in the 1990s, implying that there should have been some increase in the absolute welfare effects as well as linkage effects. All this should reduce poverty.

Instead, what can be observed is that while there has perhaps been some minor improvement in living standards in the 1990s, the overall outcome is if anything somewhat worse than the situation that prevailed in the mid-1980s, when rates of economic growth were lower. There has been a stagnation of productive employment opportunities, despite the slightly increased rate of growth.

THIS leads to a number of questions. What explains this persistence of poverty and slow growth of employment generation even in a period of higher income growth overall? And why has this been associated even with some degree of employment diversification in the rural areas in the 1990s? What has happened to the substantial inflow of resources into the rural areas and why have they been relatively ineffective? Why has domestic manufacturing industry not grown faster? What pattern of economic growth could have generated this peculiar outcome? The answers to these questions may perhaps be found in two simultaneous macroeconomic processes that are likely to have been operating in Bangladesh over the 1990s. One, which is associated with the effects of economic liberalisation, had depressing effects on real economic activity. The other, associated with the increased flow of resources to rural areas, is a more positive process, which is associated with rural employment diversification.

One of the chief aspects of economic liberalisation in Bangladesh relates to its impact on agriculture. The process of trade liberalisation covered both industry and agriculture, and especially for agricultural commodities there was a definite move towards the withdrawal of the state in importing activity. In the critical area of import of food, for example, which was completely a government activity, private grain import was permitted in 1992 and thereafter has grown substantially to the point where private traders are estimated to account for nearly 60 per cent of all food imports.

The economic reforms of the 1990s also involved a dramatic reduction - indeed elimination - of subsidies for agricultural production. Fertilizer subsidies were withdrawn; electricity and water rates for cultivation were hiked; various other subsidies were reduced and the availability of priority sector credit was substantially reduced. All this meant a substantial increase in agricultural costs.

Meanwhile, the freeing of imports meant that cheaper rice could be imported (from West Bengal in particular, but also from Vietnam and Thailand) and this meant that agricultural prices did not rise very much, and certainly rose less than proportionately to cultivation costs.

There was stagnation in food output for most of the decade, and dependence on imports increased. Per capita availability of foodgrain fell continuously during 1991 to 1996, and only increased again in 1997-98, but still not enough to regain the level achieved at the beginning of the decade. It is estimated to have fallen again since then. So we have the extraordinary combination of falling prices, less per capita availability and higher costs in food production.

Since government operations in procurement were very limited, the procurement price could not play a role as minimum support price for farmers. As a consequence, terms of trade moved against agriculture over the decade, mirroring the pattern of international relative prices. But in Bangladesh this occurred even as costs of cultivation were rising quite rapidly. The result was to create a major disincentive for agricultural investment, which indeed has slumped over the decade.

Low investment in turn leads to continued low productivity, which contributes to the spiral of higher per unit costs. Thus food production (and agriculture in general) has been affected by the vicious negative spiral, which involved prices of food remaining low in terms of other prices in Bangladesh, but high (at prevailing nominal exchange rates) relative to other competing manufacturing countries in the region. This in turn has meant that real wages of workers in Bangladesh have been lower, but nominal wages (once again, at prevailing nominal exchange rates) have been higher than those in, say, India - or rather West Bengal.

Of course, higher nominal wages mean higher nominal wage costs in all production, not just in agriculture. So the policy of liberalising agricultural trade and reducing/eliminating subsidies had an adverse impact on industrial production as well. This tendency towards higher costs is apparent in labour-intensive manufacturing, which should be very much a part of the Bangladesh economy's comparative advantage. Thus high prices of agricultural inputs and food, as well as high prices of certain imported inputs and a reduction of various subsidies such as for energy use, in manufacturing and so on, have greatly increased costs in this sector.

Meanwhile, the industrial sector in turn has been negatively affected by import competition, as evidenced both by the high degree of officially declared import penetration as well as the large unofficial estimates of smuggling of manufactured goods across the border from India. The diversion of domestic demand towards externally produced manufactured goods (whether through imports or smuggling) has meant that domestic demand for domestic production remained low or fell.

All this acted as a major disincentive for increased investment, which in turn helps to explain why the investment rate remained relatively low; why there is evident de-industrialisation for certain regions and especially certain sub-sectors; and why employment generation in manufacturing has remained so low. Also, the existence of increasing returns to scale in some manufacturing industries (such as chemicals) has compounded the problem, since lower capacity utilisation means an increase in per unit costs.

THIS points to some important policy considerations that have relevance for other developing countries, including India: the possible negative effects of subsidy removal for costs in both agriculture and industry; the effects on costs of both the pattern and the sequencing of import liberalisation; the effects of low domestic demand in terms of raising per unit costs, thus depressing investment and perpetuating low productivity.

However, it was not only this rather depressing process which was in operation in Bangladesh in the 1990s. A more positive and dynamic macro-economic process was also in operation, but this had less to do with market-orientation and more to do with the activities of the state and the NGOs. Thus, the 1990s witnessed coterminous tendencies which affected especially the rural areas:

1. The expansion of rural transport infrastructure, especially roads, bridges, bus links and so on. This reflected the substantial increase in public expenditure directed to these areas, and in turn generated large linkage effects in terms of greater access to markets, information and so on.

2. The increase in other social infrastructure in the rural areas, especially in primary education. This contributed to the diversification of employment into non-agriculture, not only because of the requirements of labour from such expenditure, but also because of the fact that some degree of education tends to encourage employment diversification.

3. The continued growth of micro-credit programmes, largely NGO-based, which increased by many multiples in terms of both spread and resource provision over this period. For many of the NGOs, the amounts provided as loan can no longer be called "micro" because they cover loans up to one lakh taka for periods up to one year. While the macroeconomic effects of these programmes may not be very apparent, several micro-studies have highlighted many positive effects. Certainly they must have some linkage effects and have contributed to rural employment diversification into non-agricultural areas. But they also have very significant gender effects, since so much of the credit is directed primarily towards women, and it has been argued that this is one very important factor that explains the decline in the fertility rate in Bangladesh over the decade.

These tendencies together have created some degree of dynamism in the Bangladeshi countryside which may not be entirely captured by the official data. In particular, they have contributed to the expansion of non-agricultural employment in the rural areas, not all of which is simply distress movement out of agricultural employment.

This may also help explain the relative resilience of the living standards of the people in the wake of the most devastating floods ever experienced in living memory, unsurpassed in terms of the duration as well as the amount of land area submerged. It is too soon to know the full effect of the floods in terms of the actual effects on increased poverty and other problems. But a rapid assessment undertaken by the Bangladesh Institute of Development Studies has found that while the effects are undoubtedly severe, they are perhaps less extreme than was predicted, and possibly less than the effects of the 1988 flood, which was actually less extensive in terms of physical damage.

This positive process, however, does not mean that there is any scope for complacency about the future of the Bangladesh economy. Rather, given the problematic macro-economic scenario outlined earlier, it suggests that even the public and NGO expenditures which have made such dynamism and resilience possible, may become unsustainable unless the roots of the macro-economic problem are dealt with. This effectively means pushing up domestic savings and investment rates, providing some degree of protection to domestic production and working out ways of feasible cost reduction, including public provision of some subsidies.

In effect, this means a reversal of some of the policies instituted earlier in the decade. Unfortunately, as long as economic policy in Bangladesh remains so donor-driven and so bound by the straitjacket of the now-discredited "Washington consensus" , it is difficult to see how this can come about.