External liberalization, economic performance, and social policy, J Berg, L Taylor

Tags: liberalization, increase, Employment, Income Inequality, social policy, Real exchange rate, Growth rate, Interest Rate, Aggregate Demand, Participation Rate Unemployment Rate, Employment rate, government spending, labor market, sector, productivity growth, effective demand, formal sector, Russia, wage inequality, labor productivity, financial crisis, Turkey, social policies, financial liberalization, exchange rate, labor demand, unskilled labor, financial sector, Current account, sectors, production factor, employment levels, government revenues, rural development, social services, government, social expenditures, employment opportunities, savings rate, social security coverage, Private sector employment, skilled labor, capital account, exchange rates, capital accounts, domestic financial markets, Economic Performance, effects of globalization, Latin America, Unemployment Rate, Lance Taylor, interest rates, capital inflows, traded goods, middle income countries, Economic Policy Analysis, macroeconomic model, labor markets, Mexico, trade liberalization
Content: CEPA Center for Economic Policy Analysis External Liberalization, Economic Performance, and Social Policy Janine Berg and Lance Taylor (CEPA) CEPA Working Paper Series I Globalization, Labor Markets, and Social Policy A Project Funded by the John D. and Catherine T. MacArthur Foundation Working Paper No. 12 February 2000 Center for Economic Policy Analysis New School University 80 Fifth Avenue, Fifth Floor, New York, NY 10011-8002 Tel. 212.229.5901 Fax 212.229.5903 www.newschool.edu/cepa
1 14/2/00 External Liberalization, Economic Performance, and Social Policy by Janine Berg and Lance Taylor As seen from the year 2000, economic policy in developing and post-socialist economies during the preceding 10 to 15 years had one dominating theme. Packages aimed at liberalizing the Balance of Payments, on both current and capital accounts, showed up throughout Latin America, Eastern Europe, Asia, and even in parts of Africa. Together with large but highly volatile foreign capital movements (often but not always in connection with privatization of state enterprises), this wave of external deregulation was the central feature of "globalization" for the non-industrialized world. This volume draws together recent historical evidence from nine developing and transition economies - Argentina, Colombia, Cuba, India, South Korea (hereafter simply referred to as "Korea"), Mexico, Russia, Turkey, and Zimbabwe - to study and assess these changes and their economic and social consequences. A companion book edited by Ganuza, Taylor, and Vos (2000) covers the experience of the four Latin American countries just mentioned, along with ten others. The discussion here concentrates on the sample of nine countries, with results from Latin America brought in where they add information. This analytical introduction summarizes the results of the studies from several perspectives. It begins with a review of different ways of analyzing liberalization and globalization that appear in the literature. Then the basic approach of the country papers and their key results regarding growth and inequality are quickly presented. These points are elaborated in the form of a simple macroeconomic model that captures the flavor of liberalization on the ground. An initial application of the model is followed by a review of decomposition exercises for effective demand and employment and distributional changes that are presented in the papers. The results of the decompositions and other salient indicators of the effects of liberalization are displayed in a set of tables, which are used to construct vignettes summarizing the country experiences. Next comes a discussion of the social policy issues emphasized by the country authors. The chapter closes with
2 a review of open questions regarding macroeconomic and external policy alternatives, and whether social policy can ameliorate the worst effects of globalization and balance of payments deregulation. Views about Liberalization Liberalization arrived abruptly. Stabilization and structural adjustment efforts through the mid-1980s had concentrated on fiscal and monetary restraint and realignment of exchange rates. Then in the late 1980s and early 1990s came drastic reductions in trade restrictions and domestic and external financial liberalization, almost simultaneously in most countries. Complementary steps were also taken toward restructuring domestic financial markets, tax systems, and labor markets. All these changes are very recent. It will take time before their full effects on growth, employment, income distribution, and poverty can be fully assessed. But external liberalization marks a dramatic switch in development policy away from traditional regimes of widespread state controls and import-substituting industrialization. One would expect to see large consequences. The old policy model had been criticized for failing to promote efficient and competitive industrial production, for creating insufficient employment, and for failing to reduce income inequality. Its rapid abolition raises a new set of fundamental questions. Will the liberalization of trade and capital flows help countries meet social goals such as reductions in inequality and poverty, better provision of health and education, and social security? Will a world system in which national economies are highly integrated in commodity and capital markets (in terms of both increased transactions flows and tendencies toward price equalization) attain these goals of its own accord? Can social policies be deployed to ease the task? The main official justification for the reforms was stated in terms of visible increases in economic efficiency and output growth that they were supposed to bring. Governments and International institutions promoting them were less explicit about their distributional consequences. The predominant view is that liberalization is likely to lead to better economic performance, at least in the medium to long run. Even if there are adverse transitional impacts,
3 they can be cushioned by social policies, and in any case after some time they will be outweighed by more rapid income growth. This conclusion is fundamentally based on supply-side arguments. The purpose of trade reform is to switch production from non-tradable goods and inefficient import-substitutes towards exportable goods in which poor countries should have comparative advantage. Presumed full employment of all resources (labor included) enables such a switch to be made painlessly. Opening the capital account is supposed to bring financial inflows that will stimulate investment and productivity growth. In a typical mainstream syllogism, Londoсo and Szekely (1998) postulate that equity is positively related to growth and investment. These in turn are asserted to be positively related to structural reforms, so the conclusion is that liberalization supports low-income groups. A second position is more radical in that its proponents such as Rodrik (1998) and Sen (1999) argue that social policies should be deployed to help the poorest, on the implicit assumption that the forces determining the income distribution, the extent of poverty, and social relationships more generally are largely independent of liberalization and globalization. Finally, others argue that while there may be supply-side benefits from trade and capital market reforms one should not overlook aggregate demand, its potentially unfavorable interactions with distribution, and the impact of capital inflows on relative prices. The importsubstitution model relied on expansion of internal markets with rising real wages as part of the strategy. Under the new regime controlling wage costs has come to center stage. So long as there is enough productivity growth and no substantial displacement of workers, wage restraint need not be a problem because output expansion could create space for real income growth. But if wage levels are seriously reduced and/or workers with high consumption propensities lose their jobs, contraction of domestic demand could cut labor income in sectors that produce for the local market. Income inequality could rise if displaced unskilled workers end up in informal service sector activities for which there is a declining demand. Rising capital inflows following liberalization tend to lead to real exchange rate appreciation, offsetting liberalization's incentives for traded goods production and forcing greater
4 reductions in real wage costs. Appreciation in turn may be linked to high real interest rates, which add to production costs and penalize capital formation. Higher rates may also draw in more external capital, setting off a high interest rate/strong exchange rate spiral. Via the banking system, capital inflows feed into international reserves and domestic credit expansion. On the positive side, more available credit may stimulate aggregate spending through increased domestic investment. However, credit expansion can also trigger a consumption boom (with the new purchases heavily weighted toward imports) or a speculative asset price bubble (typically in equity and/or real estate). The demand expansion may prove to be short-lived if the consequent widening of the external balance is unsustainable or if capital flees the economy when the bubble begins to deflate. Lack of prudential financial regulation makes the latter outcome all the more likely. The thrust of these observations is that the effects of balance of payments liberalization on growth, employment, and income distribution emerge from a complex set of forces involving both the supply and the demand sides of the economy. Income redistribution and major shifts in relative prices are endogenous to the process. Nor is social policy a panacea for rising inequality and distributional tensions. Only two countries in the sample (Korea in 1998-99 and Colombia through much of the 1990s) took advantage of strong fiscal positions to introduce large-scale programs to offset some of liberalization's adverse distributional effects. Cuba continued its long tradition of social support, and introduced new programs to offset economic dualism induced by opening the capital and current accounts. Russia, Turkey, and Zimbabwe simply lacked fiscal capacity, and cut back social programs. To a large extent, Argentina and India left pre-existing systems (with at best partial coverage of the population) intact, and Mexico abandoned policies supporting the rural poor. The bottom line is that there can be no facile conclusions about liberalization's effects, nor about how they can be contained. To date, costs in many countries have outweighed the benefits, and this situation may persist for an extended period of time.
5 The Approach of the Country Papers To a greater or lesser extent, the authors of the country studies collected in this volume adhere to the third, "structuralist" worldview mentioned above. Structuralism is not accepted in all circles. But the strength of the papers is that their shared analytical stance eases the task of cross-country comparisons and points to coherent policy conclusions. The countries considered were selected on the basis of their economic importance and the ability of potential authors to carry high quality studies through. The sample is not large enough to be "representative," but its analysis in depth is well able to support generalizations about likely outcomes of the globalization/liberalization policy mix in other national circumstances. How did the authors separate effects of specific policy changes from other factors, such as external shocks and other policy initiatives? They addressed this standard problem in economic analysis with a mixture of the following approaches: · Well-informed country "narratives" discussing policy changes and observed outcomes in a "before-and-after" approach. The country stories started with a basic set of questions and hypotheses and a simple analytical framework suggesting possible channels of causation as outlined below. Authors sub-divided their period of analysis into "episodes" with relatively homogeneous policy packages and economic circumstances. They could then trace the effects of liberalization from one episode through another. · Still within the realm of "before and after", standardized decomposition analyses of aggregate demand, factoral income distribution, employment, and productivity growth were applied wherever data availability made them possible. These decompositions (also described below) give essential comparative information on changes in output, employment, and inequality that actually took place. · Counterfactual policy simulations ("with and without") were incorporated in some case studies, based on country-specific models. It could be argued that, ideally, an approach based on formal modeling for all countries would be a better method to verify the effects of the various exogenous shocks and policy changes. As just noted, some case studies incorporated results from such costly exercises,
6
providing useful insights. But while models may permit a more rigorous isolation of the effects of
different reform measures, they have important limitations due to their assumptions about
directions of economic causality and specification of parameters, not to mention their inability to
describe changes in behavior after liberalization and to take into account political economy. A
combination of methods is better able to provide the ingredients needed to understand the
underlying processes.
Initial Summary of Results
An immediate conclusion is that the effects of globalization and liberalization have not
been uniformly favorable. In a classification that is overly simplistic but still suggestive, outcomes
for the countries included in this study can be summarized in the following fashion:
social impacts
Favorable
Neutral
Unfavorable
Effect on growth
Positive
Argentina (until 1997-8) Mexico (post-1995)
Neutral Negative
Cuba Turkey Colombia
India Korea Mexico (pre-1995) Russia Zimbabwe
At least until 1997-98 when it fell into deep recession, Argentina was the only economy in the group with growth that accelerated for a substantial period after 1990. Mexico was a latejoining counterpart, with a rapid growth rate in the late 1990s that may not be sustainable when the United States, its largest trading partner, slips into recession. The "neutral" growth impacts in some cases are really slow-downs that can be explained by exogenous factors (e.g. Cuba's loss of support from the socialist bloc), while Russia and Zimbabwe have clearly been adversely affected by globalization. The "neutral" social impacts are in fact a mix of gains and losses, with losses dominating in the rest of the sample. Broadly similar observations apply to the Latin American case studies in Ganuza, et.al. (2000), but some country outcomes were more
7 favorable. In the 1990s, Chile, Costa Rica, the Dominican Republic, and Peru all sustained fairly high growth rates with stable or falling indexes of inequality. The somewhat illiberal policy mix that supported these "success cases" (all of which rested on sustained capital inflows) is described below. Finally, all the country histories reflect internal social and political developments that influenced economic performance. The seismic changes in Russia require no comment; although less dramatic, the breakdown of centralized planning in Korea is almost as significant. There was a strong political cycle in Turkey between redistributive and market-stabilizing policy regimes (at times under the auspices of the same government). Shifts in the other countries made fewer headlines, but may be equally as important for political economy in the medium run. A Model of Liberalization Along with the aggregate outcomes just summarized, liberalization had strong differential effects on prices and quantities in different sectors of the economy. For many but not all countries, an appropriate disaggregation of the non-financial, price/quantity side of the economy focuses on traded and non-traded goods. The key relative price is the real exchange rate or ratio of traded to non-traded goods price indexes. In more populous, less intrinsically open economies one also has to consider other price ratios such as the agricultural terms of trade (India, Turkey) or the relative price of energy products (Russia). In sub-Saharan African economies such as Zimbabwe's (not to mention primary product exporters in Latin America and the Caribbean), the terms of trade between an urban-industrial and rural-agricultural sector come to the fore. In all cases, a mixture of price and quantity adjustments to the liberalization experience is evident. Since it is broadly applicable, the traded/non-traded separation is explored in the discussion to follow. Direct effects of removing barriers to trade and capital movements show up first in the traded (or tradable) goods sector but spillovers in both directions with non-traded goods have been immediate and substantial. Amadeo and Pero (2000) and Ros (1999) point out the major connections in similar fashions. The framework is a "fix-price/flex-price" model а la Hicks (1965) and many others. Traded goods are assumed to be produced under imperfect competition. The simplest model
8 involves a discriminating monopolist manufacturing goods that can both be exported and sold at home, as in Ocampo and Taylor (1998). Households at home buy both domestically made and imported consumer goods. Prior to liberalization, firms have established mark-up rates over variable costs in both their markets - the levels will depend on the relevant elasticities. Variable cost is determined by the market prices and productivity levels of unskilled labor and intermediate imports; skilled labor and physical capital are fixed factors in the short run. The traded goods price level Pt follows from the domestic mark-up over variable cost. With stable mark-up rates, traded goods comprise a Hicksian "fix-price" sector, with a level of output X t determined by effective demand. The level of production of non-traded goods is also determined by demand, but the sector may well have decreasing returns to unskilled labor in the short run. Higher production X n is made possible by greater unskilled employment (or labor demand) Ldn . However, cost-minimizing producers will hire extra workers only at a lower real product wage w / Pn , where w is the unskilled nominal wage (fixed in the short run but subject to adjustment over time as discussed below) and Pn is the price of non-traded goods. In other words, a higher price-wage ratio Pn / w is associated with greater non-traded goods production and employment, and (if there are decreasing returns) reduced labor productivity. If Pn / w is free to vary, then non-traded goods aggregate into a "flex-price" sector. With stable mark-up rates in the traded goods sector, the inter-sectoral price ratio Pt / Pn will fall as Pn / w rises, i.e. a rising price of non-traded goods is associated with real appreciation as measured by the ratio of traded to non-traded goods price indexes (a commonly used proxy is the ratio of wholesale to retail price levels). Figure 1 gives a graphical presentation of the model.1 The key quadrant lies in the extreme northeast. It shows how prices and output in the two sectors are determined. Along the schedule for "Non-traded goods equilibrium", a higher traded goods output level X t is assumed to generate additional demand for non-traded goods. As it is met by an increase in supply, the 1 See Taylor (1991) for an algebraic treatment of linkages like those described in the text in models closely related to the one illustrated in Figure 1.
9 non-traded price-wage ratio Pn / w will rise. In the market for traded goods, depending on income effects a higher level of Pn / w can be associated with either higher or lower demand. The "Traded goods equilibrium" schedule illustrates the former case - demand for X t is stimulated by an increase in Pn / w . As drawn in the Figure, the short-run macro equilibrium defined by the intersection of the two curves is stable. FIGURE 1 HERE This equilibrium helps determine the status of several markets in the economy. For example, unskilled labor demand in the non-traded sector ( Ldn ) is determined in the northwest quadrant. Employment in the traded goods sector is shown in the second quadrant from the top on the right. A lower employment level in traded goods liberates labor that can be used in the other sector, as shown in the second quadrant from the top on the left. As the figure is drawn, labor supply Lsn exceeds demand Ldn in the non-traded sector, i.e. there is open or disguised unemployment as measured by the difference (Lsn - Ldn ) . Finally, in the extreme southeast quadrant, bigger trade deficits are associated with higher levels of X t and Pn / w . Effects of Liberalization As indicated above, in many developing economies both current and capital accounts of the balance of payments were liberalized nearly simultaneously in the late 1980s or early 1990s. Given this history, one has to consider the two policy shifts together. However, for analytical clarity it is useful to dissect them one at a time. In addition, effects of other reforms have to be considered as well, in particular domestic financial, tax, and labor market deregulation. We begin with the capital account, followed by the current account, to end with some comments regarding the other sets of reforms. Capital Account Liberalization Countries liberalized their capital accounts for several apparent reasons - to accommodate to external political pressures (Korea and many others), to find sources of finance
10 for growing fiscal deficits (Turkey, Russia), or to bring in foreign exchange to finance the imports needed to hold down prices of traded goods in exchange rate-based inflation stabilization programs (Argentina, Mexico). Whatever the rationale, when they removed restrictions on capital movements, most countries received a surge of inflows from abroad. They came in subject to the accounting restriction that an economy's net foreign asset position (total holdings of external assets minus total external liabilities) can only change gradually over time through a deficit or surplus on the current account. Hence, when external liabilities increased as foreigners acquired securities issued by national governments or firms, external assets had to jump up as well. The new assets typically showed up on the balance sheets of financial institutions, including larger international reserves of the central bank. Unless the bank made a concerted effort to "sterilize" the inflows (selling government bonds from its portfolio to "mop up liquidity," for example), they set off a domestic credit boom. In poorly regulated financial systems, there was a high risk of a classic mania-panic-crash sequence along Kindleberger (1996) lines - the famous crises in Latin America's Southern Cone around 1980 were only the first of many such disasters. When the credit expansion was allowed to work itself through, interest rates could be low. However, other factors entered to push both levels of and the spread between borrowing and lending rates upward. One source of widening spreads is related to asset price booms in housing and stock markets, which forced rates to rise on interest-bearing securities such as government debt. Another source playing a role at times originated from central banks trying to sterilize capital inflows, and so pushing up rates as well. Finally, in non-competitive financial markets, local institutions often found it easy to raise spreads. High local returns pulled more capital inflows, worsening the overall disequilibrium. Unsurprisingly, exchange rate movements complicated the story. In many countries, the exchange rate was used as a "nominal anchor" in anti-inflation programs. Its nominal level was devalued at a rate less than the rate of inflation, leading to real appreciation. In several cases, the effect was rapid, with traded goods variable costs in dollar terms jumping upward immediately after the rate was frozen.
11 The same outcome also showed up via another channel. As countries removed capital controls and adopted "floating" rates, they lost a degree of freedom in policy formulation. From standard macroeconomic theory we know that in a closed economy the market for bonds will be in equilibrium if the money market clears as well. When proper accounting restrictions (including a fixed level of net foreign assets in the short run) are imposed on portfolio choice in an open economy, this theorem continues to apply (Taylor, 1999). That is, an open economy has just one independent "asset market" relationship, say an excess supply function for bonds of the form B - B d [i, i * ,( / e)] = 0 In this equation, B and B d are bond supply and demand respectively. The latter depends positively on the domestic interest rate i , and negatively on the foreign rate i * and on expected depreciation as normalized by the current spot rate e .2 Total bond supply B will change slowly over time as new paper is issued to cover corporate and (especially) fiscal deficits. For given expectations, the formula suggests that the interest rate and spot exchange rate will be related inversely. If, for the reasons mentioned above, the domestic interest rate i tended to rise, then the exchange rate would appreciate or fall. Or, the other way `round, if the exchange rate strengthened over time, then interest rates would be pushed upward. This tendency would be amplified if real appreciation stimulated aggregate demand in the short run the other side of the coin of the well-known possibility that devaluation can be contractionary in developing economies (Krugman and Taylor, 1978). Abandoning capital controls made the exchange rate/interest rate trade-off far more difficult to manage. Some countries did succeed in keeping their exchange rates relatively weak, but they were in a minority. Summarizing, capital account liberalization combined with a boom in external inflows could easily provoke "excessive" credit expansion. Paradoxically, the credit boom could be associated with relatively high interest rates and a strong local currency. These were not the most secure foundations for liberalization of the current account, the topic we take up next. 2 Scaling the expected change in the exchange rate by its current level puts the quantity / e the expected rate of return from capital gains on foreign securities - on a comparable footing with the two interest rates.
12 Current Account Liberalization Current account deregulation basically took the form of transformation of import quota restrictions (where they were important) to tariffs, and then consolidation of tariff rates into a fairly narrow band, e.g. between zero and 20%. With a few exceptions, export subsidies were also removed. There were visible effects on the level and composition of effective demand, and on patterns of employment and labor productivity. Demand composition typically shifted in the direction of imports, especially when there was real exchange appreciation. In many cases, national savings rates also declined. This shift can partly be attributed to an increased supply of imports at low prices (increasing household spending, aided by credit expansion following financial liberalization), and partly to a profit squeeze (falling retained earnings) in industries producing traded goods. The fall in private savings sometimes was partially offset by rising government savings where fiscal policy became more restrictive. Many countries showed `stop-go' cycles in government tax and spending behavior. Especially when it went together with real appreciation, current account liberalization pushed traded goods producers toward workplace reorganization (including greater reliance on foreign outsourcing) and down-sizing. If, as assumed above, unskilled labor is an important component of variable cost, then such workers would bear the brunt of such adjustments via job losses. In other words, traded goods enterprises that stayed in operation had to cut costs by generating labor productivity growth. Depending on demand conditions, their total employment levels could easily fall. The upshot of these effects often took the form of increased inequality between groups of workers, in particular between the skilled and unskilled. This outcome is at odds with widely discussed predictions of the Stolper-Samuelson (1941) theorem, according to which trade liberalization should lead to an increase in the remuneration of the relatively abundant production factor in low and middle income countries (unskilled labor) with respect to the scarce factor (capital or skilled labor). Of course, besides considering exchange rate and capital flow effects on remunerations, the model just presented departs from the standard Heckscher-Ohlin trade theory
13 framework underlying Stolper-Samuelson by working with more than two production factors and allowing for open unemployment, factor immobility, and product market imperfections. These considerations along with changes in the sectoral composition of output, as emphasized in Figure 1, are important factors in determining the distributive effects of trade liberalization (Wood 1997). With liberalization stimulating productivity increases leading to a reduction of labor demand from modern, traded-goods production, primary income differentials widened between workers in such sectors and those employed in non-traded, informal activities (e.g. informal services) and the unemployed. Graphical Illustration of the Effects of Liberalization It is easy to trace through the implications of these changes in Figure 1, beginning with the Traded goods equilibrium schedule in the northeast quadrant. The sector was subject to several conflicting forces: · By switching demand toward imports, current account liberalization tended to reduce output X t . This demand loss was strengthened by real appreciation and weakened or even reversed by devaluation. Removal of export subsidies hurt manufacturing and raw materials sectors in some cases. · Domestic credit expansion and a falling saving rate stimulated demand for both sectors, although high interest rates may have held back spending on luxury manufactured items such as consumer durables and cars (in countries where they were produced). The outcome is that the shift in the Traded goods equilibrium schedule was ambiguous, as shown by the double-headed arrow in the diagram. The contractionary forces just mentioned did not impinge directly on non-traded goods; as shown, the corresponding market equilibrium schedule shifted upward. The likely results after both schedules adjusted were a higher nontraded price-wage ratio Pn / w , a fall in the intersectoral terms-of-trade Pt / Pn , and an ambiguous change in X t . In some cases (notably Cuba, Russia, and Zimbabwe), the increase in the "flex-price" Pn was associated with an inflationary process shifting the income distribution away from wages and toward public revenues or profits. The outcome was a reduction in effective
14 demand through "forced saving" by wage-earners with high propensities to consume, as analyzed by Keynes and contemporaries in the 1920s and Kaldor after World War II.3 Turning to employment and productivity changes, new jobs were typically created in the non-traded sector, i.e. Ldn went up along the demand schedule in the northwest quadrant. With overall decreasing returns in the sector, its real wage w / Pn and labor productivity level X n / Ldn could be expected to fall. In the traded goods sector, higher labor productivity meant that the labor demand schedule in the middle quadrant on the right moved toward the origin. Regardless of what happened to their overall level of activity, traded goods producers generated fewer jobs per unit of output. Reading through the lower quadrant on the left, Lsn or unskilled labor supply in nontraded goods tended to rise. The effect on overall unemployment (Lsn - Ldn ) was unclear. Wage dynamics appeared to be driven by institutional circumstances in partly segmented labor markets, with details differing country by country. In many cases, stable or rising unemployment and unresponsive wages caused the overall income distribution to become more concentrated. The differential between skilled and unskilled wage rates tended to rise. The final curve that shifted was the one setting the trade deficit in the extreme southeast quadrant. Higher import demand and (typically) lagging exports meant that it moved away from the origin - for a given output level, the deficit went up. The corresponding increase in "required" capital inflows fed into the shifts in the capital account discussed above. Other Reforms When assessing the effects hypothesized above in real country contexts, one has to take account of other measures that were implemented simultaneously in many places and which compounded the effects discussed above. We briefly mention three other major areas of liberalization: Domestic financial sector deregulation: the effects of capital account liberalization have to be understood in conjunction with the domestic financial sector reforms that also took place in 3 See Taylor (1991) for references and further discussion.
15 many countries before or around 1990. The lifting of interest-rate ceilings, lowering of reserve requirements, and easing of entry for new banks and other financial institutions were conducive to private credit expansion fuelled by foreign capital inflows. With inadequate bank regulation and supervision in most countries, these changes in regulatory policy exacerbated the risk of banking crises along the lines described above (Vos 1995). Labor market liberalization: Typically, only small changes have occurred in this area. However, distributional outcomes can be strongly influenced by the degree of wage rigidity and labor market segmentation. In most cases institutional wage setting in modern sector firms continues to prevail (as assumed above), as well as regulations stipulating high severance payments in case of dismissal of employees. Strongly segmented labor markets are still a main characteristic in many countries. The bargaining power of organized labor may well have declined, reducing the political space for real wage adjustments. Tax reforms: Broadly speaking, countries moved towards taxation of consumption through valued added taxes and away from direct taxation, roughly a shift away from taxing the wealthy and toward lower and middle income groups. Substantial lowering of marginal rates on income and corporate taxes has been common. Decomposition Techniques To trace through the sorts of changes described by the model in detail, the first step is to examine how major economic aggregates shift over time. To this end, the country papers deploy several simple time series decomposition techniques. The essentials are outlined in this section, beginning with effective demand and going on to employment, productivity growth, and the functional income distribution. Effective Demand Over the liberalization period, there have been substantial changes in demand-side parameters such as import coefficients and savings rates along with jumps in flows such as annual exports, investment, etc. It is illuminating to look at how output has responded to these shifts, using a simple decomposition of demand "injections" (investment, government spending, exports) versus "leakages" (saving, taxes, imports). The key point is that in macroeconomic
16
equilibrium, totals of injections and leakages must be equal. Broadly following Godley (1999) this fact can be used to set up a decomposition methodology for effective demand. At the one-sector level, aggregate supply (X) can be defined as the sum of private incomes (YP), net taxes (T) and imports (M):
(1)
X = YP + T + M
The aggregate supply and demand balance can be written as:
(2)
X = CP + IP + G + E
i.e., the sum of private consumption, private investment, government spending and exports. Leakage parameters can be defined relative to aggregate output, yielding the private savings rate as sP = (YP ­ C)/X; the import propensity as m = M/X and the tax rate as t = T/X. From this one gets a typical Keynesian income multiplier function:
(3)
X
=
sP
1 +t
+
m (IP
+G
+
E)
which can also be written as:
(4)
X=
sP
IP +
t
G +
m
E
(sP + t + m) sP (sP + t + m) t (sP + t + m) m
in which IP/sP, G/t and E/m can be interpreted as the direct "own" multiplier effects (or "stances") on output of private investment, government spending, and export injections with their overall impact scaled by the corresponding "leakages" (respectively, savings, tax, and import propensities). The country papers use equation (4) in several ways. The simplest is a diagram of stances and total supply over time. In Mexico before 1994, for example, I p / s p was substantially higher than X , as the private sector pumped demand into the system, while (E / m) < X meant that high import levels were cutting into demand. The roles of the private and foreign sectors reversed sharply after the devaluation of 1994-95. Another representation involves the levels of
17 (I p - s p X ) , (G - tX ) , and (E - mX ) which from (4) must sum to zero. Both such diagrams are helpful in identifying expansionary and contractionary factors in effective demand. Several papers apply discrete time "first differencing" techniques to (4) along the lines presented below. These show the contributions of shifting weights vs. shifting multiplier impacts in determining X . From the above equation system one can also derive the economy's real financial balance as: (5) P + Z + A = (IP - sPX ) + (G - tX) + (E - mX) = 0 where P, Z, and A stand respectively for the net change in financial claims against the private sector, in government debt, and in foreign assets. In continuous time, we have dP / dt = I p - s p X , dZ / dt = G - tX , and dA / dt = E - mX . A couple of points can be made here. First, claims against an institutional entity (the private sector, government, or rest of the world) are growing when its stance with respect to X exceeds X itself. So when E < mX , net foreign assets of the home economy are declining, while G > tX means that its government is running up debt. A contractionary stance of the rest of the world requires some other sector to be increasing liabilities or lowering assets, e.g. the public sector when G > tX . Because it is true that dP/ dt + dZ / dt + dA/ dt = 0 , such offsetting effects are unavoidable. Second, stock/flow disequilibrium problems threaten when ratios such as P / X , Z / X , or - A / X (or P / YP , Z / tX , or - A / E ) become "too large." Then the component expressions in (1) and the accumulation flows in (2) have to shift to bring the system back toward financial "stock-flow" or "stock-stock" equilibrium. Such adjustments can be quite painful. Costs associated with the accumulation of net lending over time may imply important income redistribution effects between private and public domestic agents and the rest of the world. When taking such asset-related income transfers into account, we get the more familiar macroeconomic balances linked to expenditures and savings out of the disposable income of each institution, rather than from total supply as implied by equation (5) above, i.e.,
18
(6)
Dp + Dg - (Fp +Fg ) = (IP - sPX - iDg + ei*Fp ) + (G - tX + iDg + ei*Fg)
+ (E - mX ­ ei*F) = 0
where Dp, Dg and F(=Fg+Fp) stand for, respectively, the stock of net private sector debt, net government debt, and net external liabilities, as accumulated through the financing of the three
gaps (in parentheses on the right-hand side) "after transfers" over time. The level of -F is the "after transfer" counterpart of net foreign assets A. The parameters i, i* and e in equation (6)
stand for the domestic interest rate, foreign interest rate and the nominal exchange rate. The
formula permits detailed study of shifting patterns of effective demand.
Employment Decompositions
Next, we take up decompositions of employment shifts. To save algebra, the formulas
are presented in continuous time. That is, they are not set up in terms of discrete changes of the
variables that they contain, even though this is how the data are always presented. With enough
patience in writing down discrete-time first difference expansions, the right- and left-hand sides of
all the decomposition expressions that follow can be made equal by balancing beginning- and
end-of-period terms - see Pieper (2000) for examples. Such refinement is omitted here in the
interest of ease of presentation.
In terms of notation, we consider changes from time t - 1 to t , or from time zero to time
one. The difference operator is , i.e. X = X t - X t-1 , and we set X^ = X / X t-1 to indicate a
growth rate.4 Let P be the population, E the economically active population, L the total of
people employed, and U the total unemployed or U = E - L . The participation rate is
= E / P and the unemployment rate is = U / E . The employment rate is
L / E = 1 - = / with = L / P as the employed share of the population. Evidently, we
have E = L + U Dividing by P lets this expression be rewritten as = + . Taking first
differences and a bit of algebra show that
(7)
0 = (1 - )(^ - ^) + ^ = -(1 - )^ + ^ + (1 - )^
4 An alternative notation used in the literature and some of the papers is X * = X / X t-1 .
19
The first expression basically states that changes in the rates of employment and unemployment must sum to zero. The second further decomposes this condition in terms of the participation rate , the unemployment rate , and the employed share of the population . In turn, the employment ratio, = L / P , provides a useful tool to analyze job growth across sectors. Let Li be employment in sector i , with L = Li . Let X i be real output in sector i ,
and xi = X i / P or sectoral output per capita. The labor/output ratio in sector i can be written as bi = Li / X i , and let i = Li / P . Then we have = (Li / X i )( X i / P) = bi xi . Taking
first differences gives
(8)
^ = i (x^i + b^i ) = i (x^i - ^ i )
so that the growth rate of the overall employment ratio is determined as a weighted average across sectors of differences between growth rates of output levels per capita and labor
productivity (with productivity defined as i = X i / Li , and ^ i = -b^i ). Combined with (7), equation (8) provides a framework in which sources of job creation can usefully be explored. In expanding sectors (relative to population growth), productivity increases do not necessarily translate into reduced employment; in slow-growing or shrinking sectors, higher productivity means that employment declines. Under liberalization, the interaction of non-traded and traded sectors can be traced in this fashion, along with the behavior of sectors acting as "sources" or "sinks" for labor (agriculture has played both roles recently, in different countries). Labor Productivity Growth Formalizing a suggestion by Syrquin (1986), one can also decompose growth of overall labor productivity = X / L = X i / Li . The first difference decomposition is
^ = [( Xi / X )X^ i -(Li / L)L^i ]
(9)
= (Li / L)^i + [( Xi / X ) - (Li / L)} X^ i
= ( X i / X )^i + [( X i / X ) - (Li / L)]L^i .
20
The first line decomposes overall productivity growth into movements in output and
employment, weighted by sectoral shares of these two variables. As discussed above, a common
pattern under liberalization involved slow output growth and positive productivity growth in traded
goods sectors, and faster output growth but low or negative productivity growth in non-tradeds.
Across sectors, the outcome was fairly slow productivity growth overall.
The second and third lines show how overall productivity change can be written as a
weighted average of sectoral productivity shifts plus a "correction" term involving weighted
reallocations of output or employment across sectors. The reallocation weights
[( X i / X ) - (Li / L)] reflect differing productivity levels in different sectors. An output or
employment loss in a low productivity sector (agriculture, for example, with a negative value of
[( X i / X ) - (Li / L)] ), will add to overall productivity growth, as will an employment or output gain
in a sector with a relatively high output/labour ratio. In the country studies, such reallocation
effects were observed everywhere, but were economically important in only a few cases.
Capital and Labor Productivity and Real Earnings
Assuming two labor skill or ascriptive classes, total value-added nationally or in a sector
can be written out as PX = + w1L1 + w2 L2 , where P is an output price index, w1 and w2 are wage levels for the two sorts of labor, and stands for other payment flows (profits in a
broad sense, perhaps self-employment income, etc.) Let i = wi Li / PX . The first difference
version of the decomposition of payments is then
(10)
0 = (1 - i - 2 )(^ - P^ - X^ ) + i [(w^ i - P^ ) - ( X^ - L^i )]
If a breakdown of value-added by components is available, (10) provides a useful means to think about productivity and payment shifts. If = rPK , where r is the profit rate and K the
level of capital stock, then ^ - P^ - X^ = r^ + K^ - X^ . With a rising capital/output ratio, a falling
profit rate would be needed to open room for real (product) wage growth w^ i - P^ for labor type i
to equal or exceed its productivity growth rate X^ i - L^i . In the labor market itself, moderate wage
21 and high productivity growth for skilled workers may combine with low or negative productivity and wage growth for the unskilled to maintain the equality in (10). Summary of Liberalization's Outcomes To trace through all the changes described in previous sections, the first step is to examine how major economic aggregates shifted over time. Tables 1 through 9 give overviews of the main country findings regarding growth, employment, productivity, inequality, sources of effective demand, and overall macroeconomic performance. Their periodization is based on the policy "episodes" identified by the country authors in their papers.
22
Table 1: Argentina
Growth, Employment and Inequality 1990-94
Growth rate
8.5
Real exchange rate (+=real app.)
++
Employment rate (+=fall in unemp.)
-
Wage share in GDP
--
Real Wages
+
Income Inequality
Per capita household income
-
Primary incomes (labor force)
+
Skilled/Unskilled
+
Formal/Informal
+
Employment Structure
Traded/Non-traded
-
Skilled/Unskilled
+
Formal/Informal
-
1995-96 0.4 0 -0 0 +/0 + -
1996-97 8.6 0 + 0 0 +/0 + + +/0 + -
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1990-94 9.6 + n.a. -- +/n.a. -
1995-96 0.5 n.a. + + n.a. +
1996-97 10.1 + +/0 - 0 -
Productivity and Employment Overall Productivity Growth Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Changes Participation Rate Unemployment Rate Employment Rate
1990-94 7.8 0.2 negative
1995-96 2.7 -2.0 negative
1996-97 1.2 7.3 small
+
+
+
++
++
-
--
0/-
+/0
Macroeconomic variables
1990-94 1995-96 1996-97
Trade deficit
++
-
+
Domestic credit
--
+/0
+/0
Changes in reserves
+
--
0/-
Real Interest Rate
n.a.
+/-
+/0
Interest Rate Spreads
n.a.
+/-
-
Imports/GDP
++
0
+/0
Exports/GDP
--
+
0
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,-=decrease, 0/-
=slight decrease, +/-+=fluctuating trend, +/-=up then down (or vice versa), 0=no change, n.a.=not
available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
23
Table 2: Colombia
Growth, Employment and Inequality 1992-95
Growth rate
5.7
Real exchange rate (+=real app.)
+
Employment rate (+=fall in unemp.)
0
Wage share in GDP
++
Real Wages
++
Income Inequality
Per capita household income
+
Primary incomes (labor force)
+/0
Skilled/Unskilled
++
Formal/Informal
+
Employment Structure
Traded/Non-traded
-
Skilled/Unskilled
+
Formal/Informal
0/-
1995-98 1.8 + + + 0 ++ + + 0/-
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1992-95 9.6 ++ + - ++ --
1995-98 2.1 -+/+ -
Productivity and Employment Productivity Growth Overall Traded Non-traded Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Changes Participation Rate Unemployment Rate Employment Rate
1991-95 1995-97
2.6 3.0 2.7 0.2 small
2.1 2.8 2.0 -1.5 negative
-
+
0/-
++
0
-
Macroeconomic variables
1992-95 1995-98
Trade deficit
++
0
Domestic credit
++
--
Changes in reserves
-/+
+/-/+
Real Interest Rate
++
-/+
Interest Rate Spreads
0
0
Imports/GDP
+
0/-
Exports/GDP
0
+
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,
-=decrease, 0/-=slight decrease, +/-+=fluctuating trend, +/-=up then down (or vice
versa), 0=no change, n.a.=not available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
24
Table 3: Cuba
Growth, Employment and Inequality Growth rate Real exchange rate (+=real app.) Employment rate (+=fall in unemp.) Wage share in GDP Real Wages Income Inequality Per capita household income Primary incomes (labor force) Formal/Informal Employment Structure Traded/Non-traded Formal/Informal
1989-93 -8.5 ++ +/0 -- + + + + --
1994-98 4.4 -/+ 0/+/0 ++ - --
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1989-93 -13.7 -+ + ++ + -
1994-98 7.0 ++ +/0 -0
Productivity and Employment Productivity Growth Overall Traded Non-traded Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Changes Participation Rate Unemployment Rate Employment Rate
1989-93 -8.3 -13.7 -5.0 -1.0 none +/0
1994-98 4.1 11.1 0.1 -1.7 none 0/+ 0/-
Macroeconomic variables
1989-93 1994-98
Trade deficit
--
++
Imports/GDP
--
++
Exports/GDP
--
++
Imposition of export incentives
++
++
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,
-=decrease, 0/-=slight decrease, +/-+=fluctuating trend, +/-=up then down (or vice
versa), 0=no change, n.a.=not available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
25
Table 4: India
Growth, Employment and Inequality 1986-1991
Growth rate
5.9
Real exchange rate (+=real app.)
+
Employment rate (+=fall in unemp.)
+
Wage share in GDP
-
Real Wages
+
Income Inequality
Per capita household income
+
Primary incomes (labor force)
+
Skilled/Unskilled
+
Formal/Informal
+
Employment Structure
Traded/Non-traded
-
Skilled/Unskilled
+
Formal/Informal
-
1992-1996 5.3 -/+ + - + + + -
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1986-1991 5.4 0 ++ - + -
1992-1996 7.5 + - +/--
Productivity and Employment Productivity Growth Overall Traded Non-traded Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Changes Participation Rate Unemployment Rate Employment Rate
1986-1991 1992-1996
3.8
2.5
n.a.
n.a.
n.a.
n.a.
2.0
2.0
none
negative
+
+
-
+/0
+/0
0/-
Macroeconomic variables
1986-1991 1992-1996
Trade deficit
-
++
Domestic credit
0
+
Changes in reserves
-
+
Real Interest Rate
+
+/-
Interest Rate Spreads
0
+/-
Imports/GDP
+/0
+
Exports/GDP
+
+
Imposition of export incentives
0
-
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,
-=decrease, 0/-=slight decrease, +/-+=fluctuating trend, +/-=up then down (or vice
versa), 0=no change, n.a.=not available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
26
Table 5: Korea
Growth, Employment and Inequality 1980-88
Growth rate
9.4
Real exchange rate (+=real app.)
+/-/+
Employment rate (+=fall in unemp.)
+
Wage share in GDP
+/-/+
Real Wages
6.0
Income Inequality
Per capita household income
++
Primary incomes (labor force)
++
Skilled/Unskilled
--
Formal/Informal
-
Employment Structure
Traded/Non-traded
++
Skilled/Unskilled
+/0
Formal/Informal
+
1988-93 7.2 0 ++ 9.4
1993-97 7.5 +/+/0 0/5.4
+
+/0
+
0
0/-
0
0/-
0
+
+
+/0
++
+
+
1997-98 -5.8 ----9.3 -++ ++ -++ --
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1980-88 8.3 +/0 + 0/+
1988-93 6.9 + 0 0 0/0
1993-97 9.6 ++ 0 0/- + +/0 -
Productivity and Employment Overall Productivity Growth Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Changes Participation Rate Unemployment Rate Employment Rate
1980-88 6.4 2.8 large
1988-93 4.8 2.2 large
1993-97 5.3 2.2 small
1997-98 n.a. -6.2 negative
++
++
+
-
--
0
0/-
++
+
+
+
-
Macroeconomic variables
1980-88 1988-93 1993-97 1997-98
Trade deficit
--
0/-
++
--
Domestic credit
+/0
+
++
n.a.
Changes in reserves
++
++
+
--
Real Interest Rate
+
0/-
++
+
Interest Rate Spreads
+/0
0
+/0
+
Imports/GDP
-
0/-
++
--
Exports/GDP
+
-
+
--
Imposition of export incentives
++
+
0/-
--
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,-=decrease, 0/-
=slight decrease, +/-+=fluctuating trend, +/-=up then down (or vice versa), 0=no change, n.a.=not
available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
27
Table 6: Mexico
Growth, Employment and Inequality 1985-87
Growth rate
-1.8
Real exchange rate (+=real app.)
n.a.
Employment rate (+=fall in unemp.)
-
Wage share in GDP
-
Real Wages
0/-
Income Inequality
Per capita household income
-
Primary incomes (labor force)
-
Skilled/Unskilled
+
Formal/Informal
0/-
Employment Structure
Traded/Non-traded
0/-
Skilled/Unskilled
+
Formal/Informal
-
1988-94 3.9 ++ +/0 ++ ++ + + + + -+ -
1994-95 -6.2 ---- -+ - +/0 + -
1996-98 5.8 + + 0 - + +/0 0 +/0 +/0 0 0
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1988-94 5.5 ++ +/0 - ++ 0 --
1994-95 -7.8 -0 ++ + +
1996-98 8.3 + +/0 0 0 --
Productivity and Employment Productivity Growth Overall Traded Non-traded Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Changes Participation Rate Unemployment Rate Employment Rate
1988-93 1994-97
0.6 6.0 -0.5 2.8 small
-0.8 -0.2 -2.1 3.0 small
+
+
-
++
+
-
Macroeconomic variables
1985-87 1988-94 1994-95 1996-98
Trade deficit
0
++
--
+/0
Domestic credit
+
-
+
-
Changes in reserves
-
+/-/+
--
+/0
Real Interest Rate
n.a.
n.a.
++
--
Interest Rate Spreads
n.a.
n.a.
++
--
Imports/GDP
+
++
0/-
+
Exports/GDP
+
+
++
+
Imposition of export incentives
--
0
0
0
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,-=decrease, 0/-
=slight decrease, +/-+=fluctuating trend, +/-=up then down (or vice versa), 0=no change, n.a.=not
available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
28
Table 7: Russia
Growth, Employment and Inequality 1990-92
Growth rate
-9.75
Real exchange rate (+=real app.)
-
Employment rate (+=fall in unemp.)
--
Wage share in GDP
--
Real Wages
--
Income Inequality
Per capita household income
--
Primary incomes (labor force)
--
Skilled/Unskilled
--
Formal/Informal
0
Employment Structure
Traded/Non-traded
0/-
Skilled/Unskilled
n.a.
Formal/Informal
0
1992-94 -10.7 ++ -+ +/- -+ -- +/0 --
1994-97 -2.2 ++ ---/+ +/0 0 + - -/+ +/0 -
1998 -4.6 0 - -0 0/n.a. 0
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1990-92 2.4 -++ +/0 n.a. + --
1992-94 -19.2 -+ - + ++ -
1994-97 -3.0 -+ - + ++ --
Productivity and Employment Productivity Growth Overall Traded Non-traded Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Changes Participation Rate Unemployment Rate Employment Rate
1990-92 1992-94 1994-97 1998
-7.5 -9.5 -5.5 -2.2 negative
-8.5 -11.0 -6.0 -2.5 negative
-1.0 9.0 -5.5 n.a. none
-3.0 -3.0 -4.0 n.a. negative
+/0
-
-
0
++
++
++
+
--
--
--
-
Macroeconomic variables
1990-92 1992-94 1994-97 1998
Trade deficit
+
++
+/0
-
Domestic credit
0/-
--
+
++
Changes in reserves
+
-
++
-
Real Interest Rate
--
++
+/-
--
Interest Rate Spreads
n.a.
n.a.
--
n.a.
Imports/GDP
++
--
0/-
+
Exports/GDP
++
--
0/-
+
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,-=decrease, 0/-
=slight decrease, +/-+=fluctuating trend, +/-=up then down (or vice versa), 0=no change, n.a.=not
available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
29
Table 8: Turkey
Growth, Employment and Inequality 1980-88
Growth rate
5.4
Real exchange rate (+=real app.)
--
Employment rate (+=fall in unemp.) ++
Wage share in GDP
--
Real Wages
-
Income Inequality
Per capita household income
-
Primary incomes (labor force)
--
Skilled/Unskilled
+
Formal/Informal
+
Employment Structure
Traded/Non-traded
+
Skilled/Unskilled
-
Formal/Informal
-
1989-93 4.8 ++ + ++ ++ + + ++ ++ ---
1994 -5.5 -+/0 --- + - --
1995-97 7.2 +/0 + +/0 + ++ - -
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1980-88 6.2 + -- +/-
1989-93 5.2 + ++ - + +/0 --
1994 -4.9 + - -+
1995-97 10.1 + +/0 -- + +/0 --
Productivity and Employment Productivity Growth Overall Traded Non-traded (Construction) Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Changes Participation Rate Unemployment Rate Employment Rate
1980-88* 1989-93 1994 1995-97
2.6 -2.1 8.3 3.2 small
1.7
-7.5
3.5
1.2 -13.1 3.2
2.3
-0.6
3.9
3.2
1.9
3.7
none negative none
+
+
0
+
--
-
0/-
-
++
+
+/0
+
Macroeconomic variables
1980-88 1989-93 1994 1995-97
Trade deficit
-
++
--
++
Domestic credit
+
- +/0
+
Changes in reserves
--
-/+/-
+
--
Real Interest Rate
-
+ ++
++
Interest Rate Spreads
-
+ ++
++
Imports/GDP
+
+/0 +/0
+
Exports/GDP
++
- ++
+
Imposition of export incentives
++
0
-
0
*Data for non-traded are for 1983-1998.
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,
-=decrease, 0/-=slight decrease, +/-+=fluctuating trend, +/-=up then down (or
vice versa), 0=no change, n.a.=not available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
30
Table 9: Zimbabwe
Growth, Employment and Inequality 1986-90
Growth rate
5.2
Real exchange rate (+=real app.)
-
Employment rate (+=fall in unemp.)
+
Wage share in GDP
+
Real Wages
+
Income Inequality
Per capita household income
+
Primary incomes (labor force)
+
Skilled/Unskilled
+
Formal/Informal
-
Employment Structure
Traded/Non-traded
+/0
Skilled/Unskilled
0/-
Formal/Informal
+/0
1991-92 -1.8 + 0/-- -+ + -+ -
1993-97 3.6 0 - 0/+ + +/0 -
Aggregate Demand Decomposition Aggregate Demand Direct Multiplier Effects Investment/Savings Govt/Tax Exports/Imports Effect of Leakages Savings Taxes Imports
1986-90 5.5 + + + + -
1993-97 4.6 + ++ + 0 -
Productivity and Employment Productivity Growth Overall Traded Non-traded Overall Growth in Employment Emp. Sector reallocation effects Labor Supply Decomposition Participation Rate Unemployment Rate Employment Rate
1986-90 1991-92 1993-97
1.5
-3.0
0.9
1.2
-6.8
-1.0
1.6
-0.2
2.4
2.5
1.9
1.4
none negative negative
+/0
0
+/0
-
+
0
+
0/-
0
Macroeconomic variables
1986-90 1991-92 1993-97
Trade deficit
+/0
++
-
Domestic credit
+/-/+
+
++
Changes in reserves
-
++
-
Real Interest Rate
+/-
--
++
Interest Rate Spreads
+/0
--
++
Imports/GDP
0
++
+
Exports/GDP
0/-
+
++
Imposition of export incentives
-/+
0
-/+
Key: ++=strong increase, +=increase, +/0=slight increase, --=strong decrease,
-=decrease, 0/-=slight decrease, +/-+=fluctuating trend, +/-=up then down (or
vice versa), 0=no change, n.a.=not available.
Source: Country report; World Development Indicators, 1999; ILO, 1998.
31 Growth and Equity Apart from years of overt crisis, most countries achieved moderate growth rates of GDP in the 1990s. Russia and not quite so disastrously Colombia and Zimbabwe were the main losers. Except in Korea prior to its crisis, household per capita income growth was negative or just above zero. Capital inflows increased substantially to most countries (in some cases, only prior to their respective crises). As discussed above, incoming foreign capital tended to be associated with increases in international reserves, domestic credit expansion, and real appreciation. Stronger exchange rates were generally associated with higher interest rates and increasing interest spreads. Capital inflows, credit creation, and real appreciation together stimulated aggregate demand to increase more rapidly than GDP, with consequent widening of the current account deficit. Inequality of primary incomes increased in most countries. Virtually without exception wage differentials between skilled and unskilled workers rose with liberalization, reflecting employment reallocation as suggested in Figure 1. More often than not, participation rates increased or were stable. Relative to the economically active population (following the standard definition), the unemployment rate was stable or tended to rise, again consistently with Figure 1. Sources of Effective Demand One of the principal justifications for external liberalization was its anticipated effect on trade performance. Due to efficiency gains induced by freer trade, "export-led" growth was supposed to be an immediate consequence. It did not happen, at least in terms of effective demand generation in the countries analyzed in this volume. As the country studies demonstrate, exports did tend to rise with liberalization but import leakages as well, especially when the local currency appreciated in real terms. As summarized in Tables 1-9, trade therefore held back or
32 added weakly to effective demand. The export stimulus was present, but much less strongly than originally supposed by advocates of liberalization.5 The public sector's contribution to demand varied across countries. It was positive in Columbia due to increases in social spending, Cuba as it recovered from external shocks in 1994-98, India where the consolidated government deficit has supported demand for many years, and Russia as plummeting demand was at least slowed by the fact that government spending did not decrease quite so rapidly as receipts from a failing taxation system. Elsewhere, government's impact on demand was broadly neutral. Positive or "stop-go" public sector demand effects are a surprising outcome, given the rhetoric about downsizing the state that accompanied the drive towards liberalization. Without strong contributions from the foreign and public sectors, private sector demand growth emerged as the major driving force in several of the country histories. In particular, importled consumption booms following trade and financial liberalization were the rule rather than the exception. They were triggered by both cheapening of imported traded goods (import liberalization and real exchange rate appreciation) and expansion of domestic credit supply (fomented by the surge in capital flows and domestic financial liberalization). Private savings rates fell in consequence. Fewer cases were observed in which domestic demand was driven by expanding private investment, but it did occur in Argentina and Korea early in the 1990s. The rapid reduction in demand in Russia was provoked by an investment collapse in an economy that had historically been driven by high rates of accumulation. In Mexico late in the decade, higher private capital formation could give hope for a brighter future were it not for a setback due to global instability in 1998-99. Productivity and Employment Growth With Korea prior to its crisis as a notable exception, only modest aggregate productivity increases were observed. Where data are available, they are broadly consistent with greater observed productivity growth in traded than non-traded sectors. As observed above, the change 5 By way of clarification, effects of changes in saving, tax, and import parameters are reported with positive signs in the tables. For example, the saving rate dropped sharply in Mexico in 19881994, strongly stimulating aggregate demand.
33 in aggregate productivity is result of the sum of productivity changes by sectors (weighted by sectoral output shares) plus a positive reallocation effect if labor moves from low- to highproductivity sectors. Findings from the country studies indicate that within-sector productivity shifts and output growth rates largely determined the aggregate outcomes. However, in some cases there was a negative reallocation effect as workers moved toward low productivity nontraded goods sectors. With Cuba and Russia as exceptions, the share of the economically active population (or the "participation rate") increased under liberalization. With the exception of Turkey, the unemployed as a proportion of the economically active went up as well, especially after crises and/or later in the decade. Given the modest growth of GDP noted previously, a lackluster employment performance under liberalization is scarcely surprising. Country Experiences with Growth and Distribution To complement the preceding discussion, it makes sense to consider what happened in detail in specific countries. Drawing heavily on the papers in this volume, the following sketches bring out the major points. Argentina closely fits the results of the model presented above. After liberalization and imposition of an exchange rate freeze (the "convertibility plan") in 1991, the economy continued an expansion that had started the year before. In practice, convertibility turned the central bank into a currency board. Domestic credit creation was tied directly to foreign reserves, so that the level of economic activity was driven by the volume of capital flows. In the first part of the decade, inflows were ample and the trade deficit widened. Growth continued until 1995 when the region was hit by the tequila crisis. Imports expanded rapidly until that crisis hit; exports jumped upward at mid-decade, stimulated by MERCOSUR and real appreciation in Brazil due to implementation of the Real anti-inflation plan. After the effects of the tequila shock dissipated, the economy started to grow again as capital inflows recovered, but then entered into sharp contraction in 1998-99 with slow growth and devaluation in Brazil cutting into exports.
34 Aggregate demand was led by an early investment boom and perhaps a fall in the saving rate (not estimated independently in Argentine national accounts). In labor markets, there was apparently a lagged adjustment to the wage and price realignments that took place rapidly after the exchange rate was pegged. In the industrial sector during the 1990-96 expansion, output grew at about 5% per year due to higher local consumption and exports, not fully offset by higher imports. There was rapid labor productivity growth of about 6.6% per year over the period, which forced industrial employment to fall despite the output expansion. The drop in industrial employment spilled over into other sectors, with an increase over time in open unemployment and "involuntary underemployment" (the difference between the total employment rate and the rate for full-time jobs). Per capita incomes of employed people rose until mid-decade but then declined. Between 1991 and 1998, income per capita for economically active people grew by 12.2%. Without increased unemployment and labor market restructuring the total would have been 29.4%. Labor shedding in industry also took its distributional toll. Among the economically population, the Gini coefficient rose from 0.471 in 1991 to 0.534 in 1998. By 1999, the economy was in deep recession with unemployment steadily rising. Eight years of "convertibility" ended up generating future prospects substantially worse than they were when the scheme was started. Colombia Economic performance also broadly followed the Figure 1 model. Import restrictions and export promotion schemes were reduced sharply in the early 1990s. Capital movements were liberalized, although cost-based impediments to inflows were retained. Government spending, much of it directed toward social programs, rose during the decade. The increase was largely financed by higher taxes, although the public sector did have a net expansionary effect on effective demand in the recession-plagued second half of the 1990s. Macroeconomic performance was uneven - a "go" phase in the first part of the 1990s, then a prolonged "stop." Effective demand was led by the external sector in the late 1980s. The subsequent liberalization phase (1992-95) was accompanied by a jump in private investment and a fall in saving, exchange rate appreciation, and a demand boom for non-traded goods.
35 Retrenchment began with monetary tightening in 1996. With some loosening in 1997, the monetary authorities held tight, provoking an open recession in 1998-99. One result of these changes was that the capacity of the economy to generate jobs deteriorated notably. There was labor productivity growth in both traded and non-traded sectors, in the range of 2.0-2.5 percent per year. Demand growth for traded goods was negligible throughout the decade, while non-traded demand grew at 4.5% per year during 1991-95 and 2% per year during 1995-1997. As a consequence, overall employment growth was approximately zero during 1991-95, and -1.7% during 1995-97. This poor employment performance hit workers with low levels of education the hardest. Nor did demand for their labor services appear to be sensitive to wage changes. The beneficial effects of liberalization appear to have been modest at best in terms of productivity growth, while tight money and capital inflows have been associated with an increasingly prolonged recession. Cuba The socialist model provided a basis for visible output growth through the mid1980s -- average rates were 6.5% in 1971-80, and 8.5% in 1981-85. Thereafter, a tightening external constraint and rising fiscal deficits were associated with only 0.7% growth per year between 1986 and 1989. The big external shock came between 1989 and 1993 when the import share of GDP fell from 29% to 12% while GDP per capita declined by 30%. The authorities chose to maintain both employment and nominal wages stable; as a consequence labor productivity and real wages declined in tandem with GDP per capita. With a basket of "essential" goods provided through the pre-existing rationing system, prices of freely traded commodities had to jump to force the real wage reduction. The price index for informal markets rose from 100 in 1987 to 510 in 1992 and 1553 in 1993. This inflation process was one of several in the sample involving forced saving. After 1993 there was a strong push to develop export industries, along with relaxation of state controls over external transactions and opening up of markets within Cuba for foreign exchange. Greater access to foreign exchange through tourism and other exports along with (newly liberalized) flows of direct foreign investment and remittances from abroad permitted a gradual recovery of GDP after 1993. However, growth did not exceed 2.5% per year between
36 1994 and 1998. The government was the main support for effective demand during the adjustment period, with its spending substantially exceeding revenues. The private sector was the main beneficiary; hence its effect on demand was contractionary. Exports and imports have gone up together since the early 1990s, so their demand contribution has been neutral. After 1993, productivity growth averaged 4% annually economy-wide, and 11% in traded goods sectors. These improvements took place in a rapidly evolving labor market. "Other labor incomes" (of people selling products in free markets) rose from less than 20% to 40% of GDP between 1989 and 1996, while the wage share fell from 60% to 45%. The overall profit share went up after 1993, especially for traded goods and services such as mining and tourism. There is an income gap between wage earners and more informal workers, although it has declined recently. Nevertheless, overall inequality is was higher in the late 1990s than during the 1980s, with an increase in the Gini coefficient for primary incomes from about about 0.24 to about 0.38. Provision of social services probably pushed the coefficient for secondary income down to about 0.3. India Like Argentina and Russia, India was forced into liberalization by an economic crisis in the early 1990s. There had been halting steps in a liberal direction since the 1970s, but the crisis gave orthodox reformers (both local and from the international financial institutions) a stronger hand. However, in a very populous country many segments of the economy remained delinked from international influences. Deregulation concentrated more on the current than capital account. The rupee remained unconvertible for capital transactions, a liberalization "failure" that offset the worst effects of the Asian crisis. Current account liberalization went together with dismantling an opaque and cumbersome industrial licensing system and relaxation of controls on technology imports. Despite some increase in cross-border capital movements, the exchange rate did not appreciate strongly. The real interest rate remained high (lending rates on the order of 15%), but declined a bit after the mid-1990s and spreads narrowed. Trade liberalization was associated with increases in both exports and imports; the latter increased more as a share of GDP so the trade deficit widened. The contractionary effect of trade on output was offset by the government. Indeed, a fiscal deficit has been the main support for
37 domestic demand since the 1970s. However, the fiscal position took a turn for the worse with liberalization, due to falling revenues from taxes on trade and an increasing interest burden on outstanding debt. Beginning in the mid-1990s, private investment increased, also offsetting the import drag on demand. More capital formation, however, did not translate into faster output growth. Nor did it stimulate productivity; the economy-wide rate growth rate slowed in the mid1990s. At the sectoral level, agricultural growth slowed markedly in the 1990s. The root causes were lack of rural infrastructure and the rigidity of the prevailing agrarian structure. The agricultural share of GDP declined sharply, despite a claim that liberalization would remove price biases against the sector. Since higher agricultural prices penalize the rural poor (who are net food consumers), changes in the terms of trade carry significant social and political implications. As in other countries, the labor force has experienced increased informalization. Employment in the "organized" urban manufacturing sector has declined, probably signaling job losses in household and non-factory manufacturing as well. In the early 1990s, poverty and inequality both increased, in part because of policy-induced increases in food prices and cutbacks in public expenditure. These initiatives were subsequently reversed, as policy responded to the political reaction that followed. Clearly, globalization has not served to reduce poverty and inequality. If anything, its main effect has been to slow down improvements on these fronts that had been occurring in the 1980s. Korea began to move slowly toward liberalization after a financial crisis in 1979-80. The emphasis was on the current account and domestic financial system; both had been tightly regulated by the planning bureaucracy for the preceding three decades. Pressures for liberalization came from private sector interests (consumers wanting access to more imports and business seeking more financial freedom), the international community (US corporations and financial institutions seeking access to the tightly controlled Korean market), and the fact that hands-on bureaucratic control of a complex economy had become increasingly ineffective. Economic power was concentrated in the hands of the industrial conglomerates or chaebol, complemented by a poorly understood and unregulated financial system.
38 The liberalization process accelerated sharply in the early 1990s. Cross-border financial flows were decontrolled, at the behest of the chaebol. They were seeking cheaper funds abroad and the opportunity for outgoing direct foreign investment (or DFI). Between 1993 and 1996, short-term external debt of private enterprises rose from $8 billion to $22 billion; debt of financial institutions went from $11 billion to $39 billion. Outflowing DFI jumped from $1 billion to $5 billion, mostly in holdings which became worthless with the pan-Asian crisis of 1997. The capital inflows came on the heels of a boom in the late 1980s, which had been sparked by depreciation of the won after the Plaza accord in 1985, improvements in the terms of trade (especially low oil prices), and falling international interest rates. The US forced strong appreciation in the late 1980s as its trade deficit with Korea widened. Incipient depreciation early in the decade reversed again as foreign capital moved in. Household savings rates declined in standard fashion with liberalization, strengthening the traditional pattern of private sector-led effective demand. In 1997, a collapse in export prices for memory chips, haphazard industrial policy decisions, a highly unbalanced external financial position, and lack of prudential regulation allowed the crisis in Southeast Asia to engulf Korea's far stronger economy. Recovery began in late 1998, with conflicting prospects for both the income distribution and further liberalization. Tracing the distributional effects of two decades of liberalization is not easy. Through the 1980s, unemployment decreased, the wage share increased, wage inequality (Gini coefficient and the ratio of average wages in the top and bottom deciles) declined, skill premiums fell, and the wage differential between large and small enterprises went down. Rising wage and falling profit shares put distributional pressure on the traditional growth model, which had been led by investment demand supported by high corporate and household saving rates and a fiscal surplus. A transition toward growth led by consumption from wage income is as yet incomplete. The favorable distributional trends petered out in the early 1990s, in part because of increased subcontracting by the chaebol to domestic suppliers with lower wage and productivity levels, in Korea's version of the shifts depicted in Figure 1. When the crisis hit, the IMF imposed an outlandishly intense austerity package that lasted through mid-1998. The unemployment rate rose by five percentage points and the real wage fell by 9%. Excepting the top decile which
39 benefited from higher interest rates on its assets, average household incomes fell across the board, with the greatest reductions (on the order of 20%) in the bottom deciles. Government spending on social support was increased in 1998, and following relaxation of the IMF's demand restraints there was strong output growth (partly led by domestic demand) in 1999. Whether the crisis will provoke a long-term trend toward increasing inequality in Korea remains to be seen. Mexico The liberalization process got underway in the 1980s as the economy was massively restructured after the 1982 debt crisis. It was largely complete by January 1994, when NAFTA went into effect. Financial liberalization had been completed a couple of years previously. There were two separate stages in Mexico's adjustment to liberalization, punctuated by the financial crisis of 1994-95. Beginning in the late 1980 there was a consumption boom associated with falling saving rates, rising import coefficients, and capital inflows which fed into domestic credit creation. The strong exchange rate/high interest macro price tandem was very much in evidence. This phase ended with a 45% real devaluation between 1994 and 1995. Between 1994 and 1998, exports rose from 17% to 29% of GDP, while imports went from 22% to 29%. As a consequence, the economy is now very open, with generation of effective demand led in the late 1990s by foreign trade and investment (private saving rates having recovered). The first phase featured rapid productivity growth in mining and manufactures. Agricultural employment grew at 5.8% per year during 1988-1993, while output grew at 1.9%; the sector thereby served as a labor sponge for workers displaced from the other traded goods sectors. Productivity growth in non-traded goods was -0.5% per year. In the second phase, the economy-wide rate of productivity growth fell from 0.6% per year in 1988-93 to -0.8% in 1993-97. Most sectors had negative rates, although employment growth stopped in agriculture as its productivity growth rate rose to 1.2%. While these reallocations of the labor force were taking place, the distribution of earnings became more unequal, with a rapid increase in the wage premium for skilled labor. The skill mix shifted rapidly in traded goods sectors (except agriculture) with unskilled labor serving as a component of variable cost which had to be reduced in the face of a profit squeeze due to trade
40 liberalization and a strong exchange rate. In manufacturing the wage spread grew more rapidly in subsectors in which displacement of unskilled labor was greater. As the decade closed, the economy was growing at 5-6% per year, with low levels of unemployment (that is, there was slow productivity growth). The big devaluation after 1994 clearly removed a fundamental inconsistency between liberalized markets and macro policy. Another major factor supporting Mexico's growth has been the long American upswing of the 1990s. When the US finally stops growing, the adaptability of the new Mexican model will face a major test. Russia The Russian transition doubtless has more acts to play. "Summarizing this yet unfinished drama is beyond anyone's individual capacity," the authors of the country paper rightly point out. Nonetheless, a few key scenes can be illustrated by appropriate reinterpretation of Figure 1. The "flex-price" sector in Russia in the 1990s was energy production. In Soviet times, internal energy prices were held at fantastically low levels - on the order of 10% of world prices. But by the mid-1980s, per unit output costs were beginning to rise and available supplies to fall as easily accessible oil and gas fields ran out. To allow export sales to continue, internal energy consumption had to be cut. The mechanism in wake of the Soviet collapse was a rapid rise of domestic energy prices in a generally inflationary environment. By 1997, price indexes in energyrelated sectors had gone from a value of 1 in 1991 to levels between 5 and 10 thousand; wage indexes were around 1 to 2 thousand, and other producing sectors' price indexes in the low thousands. If Figure 1's "Non-traded goods equilibrium" schedule is taken to refer to the energy sector, it shifted strongly upward during the decade. The "Traded goods equilibrium" curve can be interpreted as referring to the rest of the economy. It shifted strongly to the left as investment demand collapsed and price increases held down consumption via forced saving as in Cuba. The final outcomes were big increases in relative energy prices and reductions in most other sectors' levels of employment and economic activity. Compression of internal energy demand allowed the schedule for the trade deficit in
41 Figure 1 to remain in place or even shift toward the origin, but for reasons about to be discussed the economy at large derived no benefit. Prior to its demise, the Soviet system had two main proto-classes, the nomenklatura in charge of the party/state governing apparatus and the rest of the population. The nomenklatura were the clear gainers from the transition, as in connection with the criminal "mafia" they seized control of the major productive assets in a blatantly rigged privatization process, and engaged in massive capital flight. The capital outflow largely offset any current account improvement from higher world prices or volumes of energy exports, leaving the economy in a difficult external position. At the same time, the government's ability to collect tax and other revenues was systematically reduced, in part due to maneuvers of the ex-nomenklatura/emergent capitalist class. Fiscal spending was cut back with a lag, so that the public sector's contribution to effective demand lay above the generally falling trend line. Meanwhile, different entities in the public and private sectors built up massive totals of payments arrears with one another. This innovation certainly benefited the emergent capitalists but also helped the economy to sidestep some effects of highly contractionary monetary policy, and probably kept output reductions from being worse than they actually were. Employment even increased in relatively successfully adjusting sectors such as energy, finance and credit, and public administration, and was held fairly stable elsewhere. As in Cuba after its external shock, job protection combined with falling output and real wage reduction due to forced saving led to negative apparent productivity growth in virtually all sectors. The only Russians (the so-called "new Russians") whose real earnings rose were people in upper income strata who benefited from forced saving and the rapid, corrupt privatization. In less than a decade, the Gini coefficient literally doubled, from around 0.3 to 0.6. The rising fiscal deficit could not be financed internally. As in Turkey, the authorities turned to external borrowing to cover the fiscal gap and ratify the widening external gap (postcapital flight). Short-term government bonds were the chosen vehicle. As is well-known, their high interest rates and a spiraling stock market touched off booming capital inflows beginning at mid-
42 decade. Together with an upward blip in export prices, the foreign money gave the population some respite from the bitter initial years of the transition. But the euphoria ended with the financial crisis of August 1998. Newly rising world oil prices, increasing output in import-competing sectors stimulated by the post-crisis maxi-devaluation, and further expansion of arrears probably permitted positive GDP growth in 1999, and helped ease the worst effects of the crisis. But the income distribution remains strikingly unequal, four-fifths of the population are now poor or very poor according to the official poverty lines, overall output (even taking into account the shadow economy) is 30-40% below its level a decade previously, the production structure shows sharp duality between activities that may survive under the new economic regime and those that will not, and while Soviet-style industrial organization has been obliterated, a market-based system has not emerged in its place. At best, it will be many years before globalization and liberalization in Russia may produce happy results. Turkey was an "early reformer," with a liberalization push coming on the heels of an external crisis in the late 1970s. Developments in the 1980s and 1990s make an interesting contrast, as initial current account and labor market deregulation set up a jerky transition toward liberalized external and internal capital markets. The early 1980s witnessed a major export push, facilitated by rapid demand growth in Turkey's major trading partners and pushed on the domestic front by devaluation, aggressive export subsidies, and policies aimed at cutting real wages and the agricultural terms of trade (in contrast to India, higher agricultural prices appear to benefit - not harm - low-income peasant proprietors in the countryside). Despite rapid export growth, investment in traded goods sectors did not increase, so that capacity limits helped choke off the boom later in the decade. Moreover, higher exports were matched by imports so that demand was not externally led. More fundamentally, the model broke down as repression of wages and the terms of trade could no longer by sustained - there was a wage explosion in 1988 accompanied by a marked political shift toward "populism" а la Turk. However, the government was unwilling or unable to raise taxes to fund its higher expenditures. Liberalizing the capital account was the
43 expedient adopted to permit higher public borrowing. The pattern was for the banking system to borrow in external markets, and then re-lend the money to the government with a handsome interest rate spread. Real appreciation and a widening trade deficit led to another crisis in 1994. Real wages and the terms of trade were forced back down, along with real depreciation. But a new export boom did not follow, in part because interest rates and domestic credit creation remained high. Recovery after 1994 was led by private and public consumption, as saving and tax rates fell. In general, effective demand has not been driven by the external sector, especially in the 1990s. Productivity growth has been slow, and fairly evenly balanced between traded and nontraded goods. labor force participation has risen, accompanied by informalization and widening of wage spreads between skilled and unskilled labor. Although data are scarce, it is likely that poverty has increased. Shifts toward and away from populism on the political front were dramatic and the sequence of liberalization efforts was non-standard, but otherwise Turkey adhered quite closely to the model of Figure 1. Zimbabwe The effects of liberalization are most easily visualized if the Non-traded goods equilibrium schedule in Figure 1 is taken to refer to an rural/agricultural export sector while the Traded goods curve refers to an urban/industrial sector. Removal of protection shifts the urban curve leftward and by stimulating exports may shift the rural curve up. Spillover effects from removal of rents to import quotas previously generated in the urban sector and shifts in savings rates may complicate the effective demand and income distribution story (Rattso, 1999; Taylor, 1991). In practice, Zimbabwe entered into external liberalization in the early 1990s with an extremely closed economy. The urban/industrial sector was far stronger than in most subSaharan African economies, in part because of an import-restriction system built up in response to UN sanctions during the former Rhodesia's Unilateral Declaration of Independence period in the 1960s and 1970s. This control system was continued when the new Zimbabwen government took over in 1980. An initial boom (1980-82) was unsustainable on foreign exchange grounds. Subsequently, macro stability was maintained by using the controls to regulate the current
44 account deficit. Liberalization got underway in the early 1990s, first with deregulation of intermediate imports and then of final goods. By mid-decade the capital account had been substantially decontrolled, although some restrictions were re-imposed following currency crises in 1998. The overall liberalization experience during 1990-96 (the period covered by available numerical data) is difficult to evaluate because of a serious drought in 1992, the key transition year. The essential changes were output and investment contractions of 8-10%, a doubling of the inflation rate to 40%, a consumption boom fed by increased imports, and a tripling of the trade deficit. The latter two shifts can largely be attributed to liberalization, insofar as they replicate experience elsewhere. The inflation shock was related to several factors, including the agricultural drought, forced saving led by food prices as the supply curve shifted inward, and fiscal imbalance coming into the open. Liberalization was clearly part of the explanation. Nominal wages proved to be highly inelastic to price increases. The cut in investment could have been an accelerator response to the fall in output induced by the drought. Over the longer period, output and capital formation followed a stop-go pattern, but real GDP per capita in 1996 was below its 1992 level, and well below its gradually rising trend during the period 1985-1991. Imports and exports increased in tandem after liberalization, so external factors made a small positive contribution to effective demand. The main impetus came from reductions in the private saving rate. During 1993-97, the real exchange weakened after appreciating in 1992. Real interest rates and the interest rate spread increased, perhaps due to the need to pay higher returns on the growing public debt. Real wages and the wage share fell, consistent with the inelasticity of labor earnings to inflation noted above. Partial survey results suggest that spreads between skilled and unskilled wage rates widened. In sum, both growth and equity decreased in the wake of liberalization. This observation does not suggest that Zimbabwe should return to its previous regime of controls, since that was probably unsustainable. But it does leave open a major question about what sort of economic management can be successful in present circumstances.
45 Social Policy Following upon the preceding discussion, in this section we review the main points that the country authors make regarding social policy during the liberalization period. Argentina As already noted, the Argentine stabilization and market-friendly reforms of the 1990s produced "unfriendly" effects in the labor market. Unemployment rose from an average of 6.9% between 1990 and 1992 to 16.1% between 1995 and 1998. At the same time, the "involuntary underemployment" rate (as defined above) increased from 8.6% to 12.8%. Of the contraction in full-time employment, two-thirds of those affected were in manufacturing, two-thirds were male, and over half were heads of households. The country authors discuss social policy largely in terms of the labor market. They raise the following points: · A more flexible labor market did not alleviate unemployment, but rather contributed to underemployment and lower wages. There was a clear shift toward more intensive use of short-term contracts with lower labor costs as well as redefinition of traditional overtime work rules. · Trade liberalization concomitant with an appreciated currency accelerated the adoption of new technologies, leading to restructuring in manufacturing. Two-thirds of the contraction in full-time employment took place in that sector. (During the expansionary phase, imports satisfied more than half of increased domestic demand for manufactures.) Workers at all schooling levels were negatively affected, but job losses among the unskilled were most severe. · State-subsidized employment programs in 1997 led to a rise in female employment in health and education services. Female employment was also aided by a 1997 education reform. · Income has followed a cyclical pattern although after 1994 wage income fell while non-wage income rose. Highly-skilled employees have received the greatest increases in income. Colombia combined liberalization with an ambitious social policy program. It achieved success in terms of improvements in basic needs indicators and social service coverage, but also caused the government deficit to increase. The policy package helped cushion low-skilled
46 workers from the negative effects of trade liberalization and structural reforms. The distributional and employment effects of the globalization process included the following: · As in Argentina, liberalization and exchange rate appreciation caused severe contraction in manufacturing employment, an increase in low-skilled unemployment, and higher urban income inequality. · Declining terms of trade for the agricultural sector following from real appreciation negatively affected rural rents, hurting both landowners and (low-skilled) agricultural workers · Rising aggregate demand during the expansionary years, 1991-1995, alleviated some of the low-skilled employment loss by creating jobs in construction and transport. However, most of the employment increase was in services, benefiting higher skilled (secondary and tertiary education) workers. Overall, job creation was weak during the 1990s. · Lower labor force participation helped reduce the unemployment rate in 1993-1995, but the participation rose after 1996. By 1998, the unemployment rate had reached the highest level in recorded history (15.3% in urban areas). As noted above, significant social policy changes were implemented along with liberalization in Colombia. They included the following: · The government increased the value-added tax rate, and at the same time broadened its base. · Half of the increase in social spending went to expansion of social security coverage; the other half was transferred to regional and municipal governments for spending on education and health. · Secondary school attendance rates improved, as did coverage of water and sewerage systems, dwelling characteristics, and health indicators. · Poverty as measured by unsatisfied basic needs declined from 45.6% in 1985 to 37.2% in
47 1993. · Social spending primarily benefited poor people and rural areas. Cuba has traditionally maintained a wide range of social services, many of which continued into the globalization period. New policies were aimed at combating rising inequality between those working in sectors that generate foreign exchange and those who do not, with consequent segmentation of the labor market and income flows. The package included the following: · Legalization of remittances and holding of foreign exchange; creation of publicly owned and operated stores (including food markets) for spending foreign exchange. · Creation of public feeding services at subsidized prices for low-income people. · Maintenance of social security coverage and benefits, including health and education. · Maintaining the full-employment guarantee by not shedding all unnecessary public sector jobs; protecting those workers whose jobs were shed. · Giving state-owned farmland to cooperatives and families for agricultural production. · Allowing self-employment activities, particularly in the tourism sector. India Liberalization in the 1990s was not accompanied by an expansion of programs aimed at meeting social needs. Social policy could be characterized as "business as usual." The government maintained the package put into place in the 1970s and 1980s, without significant modifications. Some effects of liberalization are clearly of social concern: · Gains for rural workers that had been predicted by free-market reformers have not taken place. Despite an unprecedented string of favorable monsoons, the growth rate of food grain output declined in the 1990s as compared with the 1980s, implying a decrease in per capita supply. · Liberalization of agricultural prices (a policy presently being discussed) could have a devastatingly negative impact on the poor. The government probably lacks the capability to compensate the poor if the policy is put in place.
48 · There is a new fiscal squeeze due to loss in tariff revenues from liberalization; customs revenue as share of total government revenues fell from above 33% in 1991 to 23% in 1995. Public capital formation slowed notably in the 1990s. · Reforms have had an adverse effect on government spending in defense and economic services; expenditures on rural development and social (including health, education and housing) services have fallen slightly. · There is ongoing public investment on the production side of agriculture, yet it has not helped rural (especially landless) labor because non-labor inputs and capital equipment have been subsidized. Labor absorption rates in Indian agriculture are very low. · Manufacturing employment in the "formal" sector has continued a downward trend that began in late 1970s. · Private sector employment was stagnant throughout the 1990s; formal sector job growth was limited to the public sector and was positive only until 1992. There is growing informalization of the labor market. · There is no social security coverage for the 90% of the population not employed in formal, organized sector. No Western-style unemployment insurance is offered, although relief programs exist for areas suffering from drought. Some state-level programs have been notably successful. · Public expenditure on education remains low. Free and compulsory education does not exist, literacy rates are low, and there is a strong bias against primary level education. A reduction in public spending on education caused the number of primary school teachers to drop, but this trend was partially reversed after 1994. In designing social policy programs for future, it is important to keep in mind that poverty reduction in India has occurred fastest in those states that have experienced rapid rural growth with strong infrastructure development (Punjab, Haryana) and in those that relied on human resource development (Kerala).
49 Korea The 1998 financial crisis in Korea resulted in an increase in the unemployment rate to levels not seen since the 1960s, as well as a tripling of the poverty rate. However, the crisis provided momentum for expanding social expenditures. Although Korean social policy expenditures are low by international standards (even when compared to Malaysia and Thailand), the government has increased such outlays from 5% of GDP in the 1980s to 7.8% in 1997. Social changes related to liberalization and the financial crisis include the following: · After 1992, wage inequality ceased to decline. The continued rise in the share of college graduates in the labor force stopped having a positive supply-side effect on wage differentials. The financial crisis caused a 20-year trend toward lower inequality to reverse. This occurred for four reasons: (1) professional and managerial job loss was slight (2%), while 13% of production workers and laborers lost their jobs; (2) wage cuts fell most heavily on the low-skilled; (3) the construction and manufacturing sectors, where lower-paid, lessskilled labor is employed, suffered the greatest job losses; and (4) wage differentials between the chaebol and small and medium enterprises have widened. · The crisis has brought about flexibilization of the labor market; workers in large firms can no longer enjoy de facto lifetime employment. Despite the increase in social policy expenditures due to the crisis, there are pressing needs for improved and expanded efforts: · Presently most government spending goes to defense and subsidies to farmers. · Only 10% of unemployed workers received unemployment assistance in 1998. There is a pressing need to expand coverage. · There is no program aimed at helping unemployed workers find jobs. There are no programs to place workers graduating from government-established, but privately run, vocational schools. · There is a strong need to modernize tax administration and institute a more progressive tax system. · College admission procedures that rely on written examinations have encouraged the growth of private instruction to the detriment of poorer students.
50 · Health insurance, although universal, only covers limited medical expenses, typically for less serious ailments. The fee-for-service method of reimbursing doctors leads to unnecessary medical tests and rising health care costs. The government should improve monitoring of health insurance. Mexico Neither the liberalization of the late 1980s nor the financial crisis of the mid-1990s was accompanied by government efforts to improve provision of social services. Most Mexicans have seen their real wages and employment opportunities fall over the decade, leading to an increase in the incidence of extreme and moderate poverty. The income of the top 10% of the population has significantly increased. The major developments on the social policy front in the 1990s include: · Job creation has been sluggish, with employment losses in the manufacturing sector due to poor output performance coupled with productivity increases. Labor absorption has been in the low-productivity agricultural and non-traded sectors causing an overall slowdown in productivity growth for the economy, not to mention underemployment. · Due to the flexibility of the labor market, real wages have fallen dramatically, to roughly 60% of the level reached in the early 1980s. · The Gini coefficient of wage inequality has steadily increased to 0.53 in 1994 from 0.44 in 1984. High-skilled workers have benefited from increased employment opportunities and wage increases. · Rural incomes have worsened. Government "modernization" of the agriculture led to the dismantling of long-established programs aiding the sector. Farmers and farm workers have suffered from the collapse in the real guaranteed price for major crops, high interest rates, and the loss of subsidies. Russia As already noted, the political and economic transition has split the economy and society into two disparate sectors. From a quasi-egalitarian society during Soviet times, Russia today ranks among the most inequitable countries in the world, with a Gini coefficient of 0.56. The implications for social policy are profound.
51 · The transition benefited the top 10% of the population, especially high-level management of newly privatized resource-based industries and shadow-economy businesses. · Resource-based and finance industries have increased their share of employment, while employment shares in import-competing industries (machinery and metalworking, food, production materials) have decreased. Manufacturing and construction have shed 1/3 of their labor force. · There is significant unemployment for the first time since the 1930s, reaching 12% in 1998. Real average wages in the 1990s dropped by more than 50%, concomitant with a fall in the wage share of income. · There has been a severe erosion of the tax base due to privatization of state enterprises, decentralization of foreign trade and expansion of the shadow economy (presently accounting for about one-third of GDP). The Russian federal budget in 1999 was $20 billion, making it impossible to support education, Health and Social Services (in comparison, New York City's budget in 1998 was $38 billion). · Because of the erosion of the tax base, the government has had to rely on external loans, leading to greater indebtedness and concern about the availability of future resources for social spending. Without a tax base, the government has been unable to compensate its citizens for the negative effects of the transition. The end result has been significant deterioration in the living standards of the population at large. Turkey When Turkey fully liberalized in 1988, it pursued social policy goals by increasing spending on public services. However, by 1993, this effort proved to be unsustainable, leading to fiscal austerity and cutbacks of social spending. The government's inability to sustain populist policies was directly related to its failure to broaden and increase taxes on the rich. Liberalization has not had the extreme negative effects on tradable sector employment witnessed in other countries, but it has been accompanied by rising wage inequality. There has been growing informalization of the labor force. At the same time, there has also been a decline
52 in public sector employment due to privatization. Those remaining in the public sector have seen their wages rise in comparison to the private sector. Post-1993 adjustment costs fell mostly on urban workers and the peasantry. Real wages of formal sector workers dropped by 23% and rural incomes fell by 16%, both exceeding the proportionate macroeconomic costs of adjustment. The government's decision to privatize many health and education services led the private sector's share in total education and health investments to reach 50% by 1996-1997. However, this has not benefited the poor, but instead led to the emergence of elite private hospitals and schools. Zimbabwe Liberalization has hurt its previously protected industries leading to deindustrialization, as consumers have switched to imported goods. Income distribution has widened since the reforms. High-skilled workers received real wage increases while the majority of wage earners suffered significant losses in real wages due to the inflation shock of 1992. Another consequence of the reforms of the 1990s has been a tightening of government spending to the detriment of social policy. When Zimbabwe gained independence in the 1980s, it increased social policy spending and transfers in an attempt to narrow gaps between living standards of racial groups. The tax system was revised and made more progressive. Funds were used to expand the educational system and improve access to health care. Labor protection was increased and a national minimum wage was instituted, while farmers received agricultural support. For a variety of reasons, including pressures on the deficit, funding of social policies began to erode in the late 1980s. The erosion has continued into the 1990s with the following results: · User fees have been introduced for previously free education and health services. · The value of the real minimum wage has fallen with inflation. · Programs to alleviate poverty such as the 1991 Social Dimensions of Adjustment program and the 1994 Poverty Alleviation Action Plan have received inadequate funding and support, leading to scant success in reducing poverty. In sum, liberalization and adjustment have hurt the Zimbabwean poor by (1) contributing
53 to real wage and job losses and (2) forcing the government to cut social spending in an attempt to curtail the deficit during the adjustment process. Policy Alternatives The usual caveats about policy prescriptions apply. Given the diversity of country experiences just reported, it is risky to generalize about lessons and conclusions. Of course, diversity of outcomes is a result in itself. It negates general sweeping statements about whether the reforms have been exclusively beneficial or exclusively costly in terms of growth, employment, and equity. If one is to sing a sad song, however, the evidence certainly shows that in the postliberalization era few if any of the countries considered seem to have found a sustainable growth path. Employment growth has generally been slow to dismal and rising primary income disparity (in some cases over and above already high levels of inequality) has been the rule. Better performances such as those in Mexico and Korea after their financial crises (as of the year 2000, three years of sustained growth in Mexico and one in Korea) were associated with avoiding the macro price mixture of a strong real exchange rate and high domestic interest rates. Post-crisis effective demand was led by the foreign sector in Mexico and by private consumption and investment spending in Korea, suggesting that each recovering country may have its own particular demand path. Similar conclusions apply to the handful of Latin American economies described in Ganuza, et.al. (2000) that combined adequate growth with improvement or stability of indexes of inequality. Their better performances were associated with a policy mix that combined (a) avoiding a macro price mixture of real exchange rate appreciation and high domestic interest rates, (b) maintaining a system of well-directed export incentives whether put in place at the national level or as part of regional integration agreements, and (c) having a system of capital controls and prudential financial regulation able to contain the negative consequences of capital surges. For the other countries described in this volume, the news is less good. Among the historically capitalist economies, Turkey and Argentina continue to wander in a slow growth,
54 falling employment, and increasing inequality wilderness. India's growth and equity performance has not improved with liberalization, and despite a strong effort on the social policy front, Colombia's is worse. In part because of an explicit effort to cushion the liberalization shock, Cuba's equity has been maintained, despite mediocre growth performance. Zimbabwe's and especially Russia's are disasters. Of the three views regarding liberalization mentioned at the outset, the first "market friendly" narrative is hard to discern in the countries analyzed here. In line with the second view, some might argue that their distributional deterioration was not the result of liberalization and globalization but they would have to strain to make the case. For most of the countries considered here (and in the Latin American sample as well), it is difficult to refute the third view that liberalization and deteriorating growth and equity performances can easily go hand-in-hand. Finally, fundamental questions arise regarding social coherence and social policy. The mainstream view of liberalization emphasizes its likely positive effects on economic performance. Adverse transitional impacts can in principle be smoothed by social policies, and in any case after some time "a Rising Tide lifts all boats." The much more disquieting possibility is that liberalization can unleash dynamic forces leading not only to an unimpressive aggregate economic performance but also to long-term slow employment expansion and increasing income concentration. In principle, governments could put countervailing social policies into place. In practice, they probably lack the capacity to do so because of their own fiscal and administrative limitations. Such constraints on social policy and burden-sharing can be reduced by investment in the capability of the state, as experience in now industrialized countries demonstrated in the 19th century and again after World War II in the construction of welfare states (Polanyi, 1944). But an explicit political decision would be needed before such investments could be undertaken. It would be comparable in scope to the one that led to the worldwide spread of liberalization in the first place. Nevertheless, for the countries considered here, the initial outcomes of liberalization suggest that a "double movement" б la Polanyi, first toward and then away from an extreme liberal policy stance, could be forthcoming in the not-so-distant future. Inadequate social
55 performance of any economic policy line leads ultimately to its reversal as society organizes to protect its own.
56 References Amadeo, Edward J., and Valeria Pero (2000) "Adjustment, Stabilization, and the Structure of Employment in Brazil," Journal of Development Studies, forthcoming Ganuza, Enrique, Lance Taylor, and Rob Vos (2000) Godley, Wynne (1999) "Seven Unsustainable Processes," Annandale-on-Hudon, NY: Jerome Levy Economics Institute Hicks, John R. (1965) Capital and Growth, Oxford: Clarendon Press. Kindleberger, Charles P. (1996) Manias, Panics and Crashes (3rd Edition), New York: John Wiley and Sons. Krugman, Paul, and Lance Taylor (1978) "Contractionary Effects of Devaluation," Journal of International Economics, 8: 445-456 Londoсo, Juan Luis and Miguel Szйkely (1998), "Sorpresas Distributivas despuйs de una Dйcada de Reformas," Pensamiento Iberoamericano-Revista de Econуmica Polнtica (Special Issue). Ocampo, Josй Antonio and Lance Taylor (1998), "Trade Liberalization in Developing Economies: Modest Benefits but Problems with Productivity Growth, Macro Prices and Income Distribution," Economic Journal, 108: 1523-1546. Pieper, Ute (2000) "Openness and Structural Dynamics of Productivity and Employment in developing countries: A Case of Deindustrialization?" Journal of Development Studies, forthcoming Polanyi, Karl (1944) The Great Transformation, New York (NY): Rinehart Rattso, Jorn (1999) "Income Distribution, Growth and Protectionism in Sub-Saharan Africa and the Case of Zimbabwe," Rodrik, Dani, (1998) "Who Needs Capital Account Convertibility?" Harvard University, Cambridge (MA) (processed). Ros, Jaime, (1999), "La Liberalization de la Balanza de Pagos en Mexico: Efectos en el Crecimiento, el Empleo y la Desigualdad Salarial," Paper presented at UNDP-CEPAL-IDB conference on the Effects of Balance of Payments Liberalization on Employment, Distribution, Poverty and Growth, Rio de Janeiro, February. Sen, Amartya (1999) Development as Freedom, New York (NY): Alfred A. Knopf Stolper, Wolfgang F., and Paul A. Samuelson (1941) "Protection and Real Wages," Review of Economic Studies, 9: 58-73 Syrquin, Moshe (1986) "Productivity Growth and Factor Reallocation," in Hollis B. Chenery, Sherwin Robinson, and Moshe Syrquin (eds.) Industrialization and Growth New York (NY): Oxford University Press Taylor, Lance (1991) Income Distribution, Inflation, and Growth, Cambridge (MA): MIT Press Taylor, Lance (1999) 'The Exchange Rate is Indeterminate in the Portfolio Balance and MundellFleming Models - Each Has One Fewer Independent Equation than People Usually Think,' New School for Social Research, New York (NY) (processed)
57 Vos, Rob (1995) `Financial Liberalization, Growth and Adjustment: Some lessons from Developing Countries', in: Stephany Griffith-Jones and Zdenek Drбbek (eds) Financial Reform in Central and Eastern Europe, London: Macmillan, pp. 179-220. Wood, Adrian (1997), "Openness and Wage Inequality in Developing Countries: the Latin American Challenge to East Asian Conventional Wisdom," World Bank Economic Review, 11: 3357.
58
Non-traded sector labor demand
Non-traded sector Lsn
L dn
labor supply
Non-traded sector pricewage ratio ( Pn / w )
Traded goods equilibrium
Non-traded goods equilibrium
Traded goods output ( X t )
Traded goods employment ( Lt )
Traded goods labor demand
Trade deficit Figure 1: Initial equilibrium positions in traded and non-traded goods markets and probable shifts after current and capital account liberalization.

J Berg, L Taylor

File: external-liberalization-economic-performance-and-social-policy.pdf
Author: J Berg, L Taylor
Published: Thu Feb 24 15:19:58 2000
Pages: 59
File size: 0.19 Mb


A compendium of evidence, 70 pages, 0.62 Mb

The question of Orientalism, 10 pages, 0.23 Mb

Removable partial dentures, 1 pages, 0.04 Mb

Evangelicals and Science, 30 pages, 0.17 Mb
Copyright © 2018 doc.uments.com