Structure of the Labor Market

Labor markets in developing countries differ in important ways from those in industrial countries. Key structural differences are the importance of the agricultural sector in economic activity (which tends to impart a marked seasonal pattern to employment in some countries or regions), the importance of self-employment, and irregular work activities. These differences imply that standard labor market concepts used in the industrial world (such as employment and unemployment) do not necessarily have the same meaning and must be interpreted with care.

Development economists typically distinguish three sectors in the labor market in developing countries (see Rosenzweig (1988)). The first is the rural sector, which is characterized by a large share of self-employed persons and unpaid family workers. The second is the informal urban sector, which emerged largely as a result of accelerated rural-urban migration and the labor surplus that it generated in the cities. The informal sector is characterized by self-employed individuals (such as small traders, street vendors, taxi drivers, tailors, carpenters, and bricklayers) or small privately owned enterprises producing mainly services and other nontradables.² Activities in this sector rely mostly on the provision of labor services by owners and their families, but occasionally also on paid labor without formal employment contracts. Job insecurity is pervasive, underemployment (as a result of low labor productivity) is high, wages are highly flexible, and workers get very few benefits from their employers. Legal minimum wage laws do not apply or are not enforced, and labor unions play a very limited role.³

The third segment of the labor market is the formal urban sector, consisting of medium and large enterprises (including state-owned firms) producing both tradable and nontradable goods using skilled and unskilled labor. Firms tend to hire workers (at least the most qualified ones) on the basis of formal contracts. Workers and employers are subject to various labor market regulations; employers, in particular, must provide a variety of benefits (such as pension plan, health insurance, and relative job security) to their workers.⁴ Labor unions and productivity considerations often play an important role in the determination of wages, and legal minimum wage laws exist—albeit enforced with varying intensity across professional occupations and across countries.⁵

The existence of different sectors in the labor market is often associated with labor market segmentation, which can be defined as a situation where observationally identical workers (that is, workers with similar productive abilities) receive different wages depending on their sector of employment. In particular, restrictions on occupational mobility between sectors— resulting from institutional barriers or otherwise—may prevent workers in the “low-wage” segment (the rural or informal sector, for instance) from having full access to a job in the “high-wage” segment (the formal sector) held by workers with similar qualifications—even if wages are fully flexible. Equilibrium of the labor market is thus characterized by the rationing of high-pay jobs, while at the same time workers able and willing to fulfil these jobs at the going wage are “stuck” in low-wage jobs. If there were no barriers, workers in the low-wage sector would enter the high-wage sector and bid wages down, leading to an equalization of sectoral earnings in the long run. While labor market segmentation exists in many developed countries, it appears to be a more pervasive phenomenon in developing countries. As further discussed below, in a segmented labor market the distinction between voluntary and involuntary unemployment lacks meaning.

The best-known model of labor market segmentation in developing nations is the migration model of Harris and Todaro (1970). The main objective of the model was to explain the persistence of rural-to-urban migration, despite the existence of widespread urban unemployment in developing countries. It suggests that migrants from the rural areas are attracted to the urban formal sector by the expectation of higher wages, even if they are unlikely to find jobs in the formal sector immediately. A key element of the model is the equality of expected (rather than actual) wages as the basic equilibrium condition across the different segments of the labor market. Specifically, Harris and Todaro assumed that rural workers, in deciding to migrate, compare (in present value terms) wages in agriculture, w_A, to the expected urban wage, $w_U^a$, which is calculated by multiplying the prevailing wage, w_u—assumed fixed as a result of the existence of, say, a minimum wage law or trade union activity—by the urban employment ratio, which measures the probability of being hired. In equilibrium, the Harris-Todaro hypothesis yields

where L_U is urban employment and N_U is the absolute number of unemployed workers in urban areas.

The Harris-Todaro model has been extended in a variety of directions over the years.⁶ A formulation similar to equation (1) has been used to explain movements of labor between the informal urban sector and the formal economy, on the assumption that congestion and relocation costs prevent instantaneous reallocation of the labor force (see Section II). Other important developments have been the explanation of urban wage rigidity as a result of the existence of trade unions and efficiency considerations internal to firms. Calvo (1978) developed one of the first models of urban wage determination based on bargaining between urban firms and a utility maximizing trade union.⁷ Models of urban wage formation based on efficiency factors were analyzed by a number of authors. According to efficiency wage theories, real wage cuts lower productivity because they directly reduce calorie intakes (Bliss and Stern (1978)) and incentives to provide effort (Stiglitz (1982)), and they increase incentives to shirk (Shapiro and Stiglitz (1984); Esfahani and Salehi-Isfahani (1989); Agénor and Aizenman (1996b)) and quits and turnover costs (Stiglitz (1974)); Agénor and Aizenman (1996a)). For instance, workers’ effort in the urban formal sector may depend positively on the wage paid in the current sector of employment relative to the wage paid in other production sectors (the informal sector wage or the agricultural wage) or the reservation wage. In such conditions, wage- setting firms will set wages so as to minimize labor costs per efficiency unit, rather than labor costs per worker. The wage that minimizes labor costs per efficiency unit is known as the efficiency wage. Firms hire labor up to the point where the marginal revenue product is equal to the wage they have set.⁸ Thus, efficiency wage theories help explain why urban- or formal- sector firms pay more than the market-clearing wage. They predict non-competitive wage differentials across segments of the labor market even in the absence of institutional impediments to wage flexibility. If, for instance, efficiency wage considerations apply differentially across sectors (owing to, say, differences in specific training costs), then intersectoral wage gaps that cannot be eliminated by market forces will emerge. Section II will examine how relative wage rigidity across skill categories induced by efficiency considerations or trade union activities alter the transmission process of macroeconomic policy shocks.

Employment Distribution and Unemployment

In many developing nations, agriculture still employs a large share of the labor force in rural areas. Over the past three decades the “modern” sector has expanded significantly in many cases (notably in Asia and part of Latin America) but continues to provide limited employment opportunities in others. The share of informal sector employment in total urban employment is sizable in many developing countries—particularly in some parts of Asia, the Middle East, and sub-Saharan Africa—and may vary between 30 and 60 percent.⁹

In Mexico, for instance, about 27 percent of the labor force is employed in agriculture, 16 percent in manufacturing, 10 percent in construction, transport and utilities, and the remainder in trade and services (Revenga (1994)). A large proportion of employment in nonagricultural sectors is found in micro enterprises and small firms, and in 1988 the informal sector accounted for about 50 percent of urban employment (Roberts (1991)). In Uruguay, informal sector activities account for 15 to 30 percent of total employment (Rama (1994b)). In India, in the mid–1980s. 70 percent of the labor force was employed in agriculture, and at least 3 out of 5 urban workers were involved in the informal sector. In Morocco, 57 percent of non-agricultural employment in the early 1980s consisted of own-account and unpaid workers and wage employment in small-scale enterprises (Said (1994)). In 1988, formal sector firms accounted for only 10 percent of all enterprises in manufacturing.¹⁰ Employment in the modern sector in Côte d’Ivoire represented about 10 percent in the early 1980s (Levy and Newman (1989)). In Tunisia, the Population and Employment Survey estimates the share of the informal sector in total nonagricultural employment at 31 percent in 1989.

The data collected by Turnham (1993) suggest the existence of an inverse relationship between income per capita and the size of the informal sector, but even in upper middle-income developing countries the informal sector (at about 30 percent) continues to account for a sizable part of total employment. There is also evidence that the size of the informal sector has increased in the past few years. For instance, the share of the informal sector in nonagricultural employment in Latin America rose from 40 percent in 1980 to 47 percent in 1985 and 55 percent in 1993.¹¹ At the same time, the share of the modern sector (medium and large private enterprises and the public sector) in total employment fell from about 60 percent in the early 1980s to around 45 percent in 1993.

As a result of the importance of the rural and urban informal sectors, the proportion of wage earners in total employment in developing countries tends to be much lower than in the industrial world, although large variations exist across countries and regions. Wage employment—which tends to prevail particularly in mining and manufacturing sectors, as well as in some urban services—accounts for about 10 percent of total employment in some low-income sub-Saharan African countries, but as much as 80 percent in some middle-income Latin American countries.¹² In Korea, wage workers accounted for about 60 percent of total employment in 1990 and 87 percent in mining and manufacturing, but only 8 percent in agriculture (Kim (1994)).

Public sector employment accounts for a large share of wage employment and the formal sector workforce in many developing countries. The evidence provided by Kraay and Van Rijckeghem (1995) indicates that government employment (central government only) as a share of formal sector employment was 23 percent between 1972 and 1980, and 28 percent between 1981 and 1992 (28 and 35 percent respectively in sub-Saharan Africa).¹³ The ratio of employment in the consolidated public sector over total formal employment during 1972–87 amounted to 21 percent in Malawi, 63 percent in Gabon, and 68 percent in India. In Egypt, in 1990, 57 percent of all urban wage earners were employed in the public sector, with 37 percent in the government and 20 percent in the public enterprise sector. The public sector has also been the dominant employer of educated labor in that country. 74 percent of urban workers holding an intermediate or higher degree were employed in the public sector in the late 1980s (Hollister and Goldstein (1994), p. 30). In Jordan, 40 percent of the population of formal sector workers are employed by the government (Hollister and Goldstein, p. 32). In Tunisia, the public sector (which includes both employment in parastatal enterprises and regular government services) accounts for about 25 percent of total employment (Hollister and Goldstein, p. 33). More than 80 percent of educated labor, and 40 percent of all wage earners in rural areas, were employed in the public sector in the late 1980s.

Available information on the distribution of public sector employment across different levels of government is relatively scant. The data compiled by Kraay and Van Rijckeghem (1995) for a group of nine developing countries indicate that on average central government employment accounts for 44 percent of total public sector employment; local government, 29 percent; and public enterprise employment, 25 percent. There are, however, substantial variations across countries and regions, which are partly related to the degree of government centralization and the degree of government involvement in “strategic” industries. Said (1994) estimates that employment in public enterprises, measured as a percentage of nonagricultural employment, stood at about 14.5 percent in Arab countries in the early 1990s—compared to less than 6 percent for Latin America, 16 percent for Asia, and 19 percent in Africa, in the early 1980s.

During the early 1980s, employment in the public sector grew faster than in the private sector in many countries of Latin America and especially Africa (Lindauer, Meesook, and Suebsaeng (1988)). The expansion in public sector employment was in part a response to adverse conditions in private labor markets—in addition to growing demand for public services, such as education and health—sometimes giving governments the role of “employer of last resort,” especially of those people with higher education levels.¹⁴ In recent years, in Algeria, Jordan, and Tunisia, government employment played a countercyclical role—rising significantly during periods of weak economic activity (Said (1994), p. 3). Public sector employment may provide a variety of benefits that help attract workers: relative job security and sometimes less than complete enforcement of performance standards, nonwage entitlements (such as subsidized housing), enhanced social status, and opportunities for moonlighting and rent-earning in some government positions (Gelb, Knight, and Sabot (1991)). Stevenson (1992) has argued that during the 1970s, the attraction of public sector jobs resulted mostly from relatively higher wages, whereas during the 1980s (a period during which real public sector wages fell in many countries, as discussed below) the attraction was related more to job security, fringe benefits, and the possibility of moonlighting. The combination of attractive public sector jobs and government hiring policies may be an important source of “wait” unemployment, as argued, for instance, by Dickens and Lang (1995) in the case of Sri Lanka. Public sector employment may also be inefficient and unproductive, and the cost in terms of forgone income may be high (Turnham (1993)).¹⁵

Published data on unemployment in developing countries are not very reliable and often incomplete. They usually include unemployed workers looking for jobs in the formal sector, but not underemployed workers in the informal and rural sectors—thus understating the effective excess supply of labor—and do not account for the fact that job seekers may be employed part of the time in the informal sector. Very few countries provide information on the duration of unemployment. Nevertheless, available data suggest that open unemployment is concentrated in urban areas and is mostly associated with wage employment, and that underemployment is far more pervasive than open unemployment.¹⁶ In Ghana, for instance, in 1988–89 the officially reported unemployment rate was 1.6 percent of the labor force, but underemployment was 24.1 percent (World Bank (1995), p. 28). In some countries, open and disguised unemployment combined amount to as much as 60 percent of the labor force (Turnham (1993)).

Figure 1 shows the behavior of output growth and the open unemployment rate for several developing countries over the first and second part of the 1980s. The data indicate a very weak correlation between the rate of output growth and the unemployment rate. Several authors have argued that the absence of a stable output-unemployment relationship may be the result of spillover effects across different segments of the labor market and shifts in production activities, which are not properly accounted for in published employment and output data. Turnham (1993), for instance, suggests that during the recession of the 1980s, the loss of jobs in the formal or modern sector led to a sharp increase in self-employment in several Latin American countries. In Argentina and Brazil, in particular, the urban informal sector was a major absorber of the increased labor supply during the 1980s, while the open unemployment rate failed to increase.¹⁷ During the recession of 1981–83 in Brazil, formal private sector employment declined sharply but employment in the informal sector increased significantly—thus mitigating the increase in open unemployment. As documented by Urani and Winograd (1994), the increase in the rate of output growth between 1983 and 1986 was also accompanied by a reduction in the share of informal sector employment, whereas the slowdown recorded between 1987 and 1988 (following the collapse of the Cruzado Plan) led to an increase in that share. In Africa informal sector employment also expanded sharply during the late 1970s and early 1980s (notably in Kenya, Zambia, and Nigeria), thus moderating increases in open unemployment, as well as the fall in output. The analytical framework examined below captures the interactions between the formal and informal sectors, as well as the mechanisms through which the adverse output and employment effects of macroeconomic shocks are mitigated by a shift to informal production activities.

Output Growth and Unemployment in Developing Countries

(Period averages, in percent)

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Sources: ILO, Yearbook of Labor Statistics: and IMF, International Financial Statistics.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Output Growth and Unemployment in Developing Countries

(Period averages, in percent)

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Sources: ILO, Yearbook of Labor Statistics: and IMF, International Financial Statistics.

• Download Figure
• Download figure as PowerPoint slide

Output Growth and Unemployment in Developing Countries

(Period averages, in percent)

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Sources: ILO, Yearbook of Labor Statistics: and IMF, International Financial Statistics.

• Download Figure
• Download figure as PowerPoint slide

The composition of unemployment by skill categories varies considerably across countries. As documented by Riveros (1990), in Latin America—particularly in Colombia, Uruguay, Argentina, and to a lesser extent in Chile and Brazil—a large proportion of the unemployed were skilled workers in the 1980s. Evidence supporting the view that skilled workers often do not work in the informal sector can be found in a detailed study of industrial workers in Brazil by Hirata and Humphrey (1991), who show that skilled workers were more likely than other labor categories to remain in open unemployment, rather than working in the informal sector. In India, the open unemployment rate is also higher among the educated (Banerjee and Bucci (1995)). Hollister and Goldstein ((1994), p. 65) provide evidence of high levels of skilled labor unemployment in several countries in the Near East—particularly in Egypt, where a large majority of those in open unemployment have secondary or postsecondary degrees. The evidence reviewed by Said (1994) also suggests that unemployment in several Arab countries is quite high for workers with higher levels of education.

Skilled workers often choose to remain unemployed, because they do not want to accept low-income jobs in the informal sector. They are thus, in a sense, “quasi-voluntarily” unemployed. This may be because their reservation wage is higher than the going wage in the informal sector, because job search in the formal sector is more efficient while unemployed, and/or because the higher family income of the educated allows them to remain unemployed while searching for a job (Banerjee and Bucci (1995)). Unskilled workers, by contrast, often cannot afford to remain unemployed for long and are often “forced” to enter the informal sector. Unskilled unemployment may nevertheless emerge if (as in the Harris-Todaro framework presented above) workers who expect to be hired at the higher formal sector wage are able to wait for the good jobs. The macroeconomic model discussed in Section II will integrate both “quasi-voluntary” or “luxury” unemployment of skilled workers and “wait” unemployment of unskilled workers.

Labor Market Institutions and Regulations

Allocation of the workforce and wage formation depend critically on labor market institutions and government regulations. As argued earlier, trade union activity and minimum wage laws may represent important sources of labor market segmentation. These and other institutional features of the labor market—such as wage indexation, labor tenure laws, and (in some countries) restrictions on labor mobility aimed at influencing migration flows—have been blamed for pushing labor costs above market- determined levels, for contributing to larger disparities in the ratios of wages to the marginal product of labor (particularly in urban areas), and for limiting the ability of firms to adjust production patterns to changes in relative prices, factor supply, and aggregate demand conditions.

In what follows we describe the main features of labor market institutions and regulations in developing countries, and will analyze in a more detailed manner in Section III their effects on the functioning of the labor market. We begin by examining hiring and firing regulations, and then turn our attention to minimum wage laws, nonwage labor costs and unemployment benefits, indexation practices, and bargaining structures.

Minimum Wage Laws

The effect of minimum wages on the labor market can be illustrated with a simple graphical analysis, in the presence of perfect and imperfect labor mobility. Consider a small open developing economy in which the formal sector produces a traded good whereas the informal sector produces a non-traded good. Both goods are produced using only homogeneous labor, the total supply of which is given. The determination of wages and employment under four different assumptions regarding labor market adjustment is shown in Figure 2. In all four panels of the figure the horizontal axis measures total labor available in the economy, O_FO_I. The vertical axis measures the wage rate in the economy, which is either uniform across sectors or sector specific. The demand for labor in the formal (informal) sector is represented by the downward-sloping curve $L_F^d\left(L_1^d\right)$Consider first panel A, which assumes that wages are perfectly flexible and labor perfectly mobile across sectors. The initial equilibrium position of the labor market obtains at point E, where the economy-wide wage rate is equal to w*, labor employed in the formal sector is O_FL_F, and labor used in the informal sector is $L_F^{*}{O}_1$

Minimum Wages, Labor Mobility, and Adjustment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Minimum Wages, Labor Mobility, and Adjustment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

Minimum Wages, Labor Mobility, and Adjustment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

Panels B, C, and D assume that the wage rate in the formal sector is fixed at the minimum level $w_F^c$(above the economy-wide, market-clearing wage) while wages in the informal sector remain flexible.²⁰ However, the panels differ in the underlying assumptions regarding the degree of intersectoral labor mobility. In panel B, labor can move freely across sectors, as in panel A. Perfect labor mobility, together with wage flexibility in the informal sector, prevents the emergence of unemployment. The initial equilibrium obtains at point A in the formal sector and corresponds to an employment level of $O_F{L}_F^c$, and at point E_I in the informal sector, with wages equal to w_I and employment to $L_F^c{O}_I$. In panel C, labor is completely immobile. The labor force in the formal sector is equal to $O_F{\overline{L}}_F$, while the supply of labor in the informal sector is measured by ${\overline{L}}_F{O}_I$. Since sectoral labor supply is completely inelastic and minimum wages are binding in the formal sector, unemployment will typically emerge in that sector. The situation depicted in panel C shows that employment in the formal sector is equal to $O_F{L}_F^c$and unemployment to $L_F{L}_F^c$. Finally, panel D illustrates the Harris-Todaro labor allocation mechanism, which assumes that equilibrium obtains when the wage rate in the informal economy is equal to the expected wage in the formal sector. In turn, as discussed above, the expected wage in the formal sector is defined as the product of the actual wage in that sector (that is, the minimum wage) times the probability of being hired, which is measured by the employment ratio. The downward-sloping locus QQ is a rectangular hyperbola along which the above equality holds, and is known as the Harris-Todaro curve (Corden and Findlay (1975)),²¹ The intersection of the $L_I^d$curve with QQ determines the wage rate and the employment level in the informal sector, while the intersection of the $L_F^d$curve with the horizontal line drawn at $w_F^c$ determines employment in the formal sector. The initial equilibrium is therefore characterized by unemployment, which is equal to $L_F^c{L}_I$

Suppose that as a result of a macroeconomic or structural shock—a permanent reduction in. say, autonomous demand—the demand for labor in the formal sector falls, shifting the curve $L_F^d$to the left.²² If wages are perfectly flexible and labor perfectly mobile across sectors, adjustment of the labor market leads to a fall in the overall wage rate in the economy, and a reallocation of labor across sectors, leading the economy to a new equilibrium (point E^’ in panel A) with full unemployment.

Consider now what happens in the presence of a binding minimum wage in the formal sector. If labor is perfectly mobile across sectors, the demand shock leads only to a reallocation of the labor force and a fall in wages in the informal sector (panel B). However, if workers cannot move across sectors, the reduction in demand leads to an increase in unemployment in the formal sector, with no effect on wages and employment in the informal economy (panel C). With a labor allocation mechanism of the Harris-Todaro type, the demand shock reduces employment in the formal sector as in the preceding case. However, the effect on the unemployment rate is now ambiguous. This is because the QQ curve shifts to the left following the shift in $L_T^d$since the fall in employment reduces the likelihood of being hired and therefore the expected wage in the formal sector. This implies that more workers would elect to seek employment in the informal sector, bidding wages down. Employment therefore increases in the informal sector, while wages fall. However, despite the reallocation of labor across sectors, in equilibrium unemployment may well increase in the formal sector. The point, nevertheless, is that the presence of binding minimum wages in the formal sector alters the effect of exogenous shocks on wages, employment allocation, and unemployment.

In practice, the effects of minimum wages depend on both the frequency at which they are adjusted (often at the government’s discretion) and the degree to which the legislation is enforced. In an inflationary environment, the real minimum wage can fall to very low levels if it is adjusted only infrequently. In such conditions the minimum wage may not operate as a binding constraint. Excessively high minimum wages (relative to the marginal product of labor) in the presence of lax enforcement provides incentives to evade the law and operate partly illegally, or to entirely shift activities to the informal economy—in a manner very similar to a tightening of job security provisions, as discussed earlier.

Low minimum wages and lax enforcement are indeed features of the labor market in many developing countries.²³ Few countries appear to have effective minimum wage policies, as can be inferred from data on real minimum wages (compared to average wages) and the actual proportion of workers earning the minimum wage or above. In Taiwan, minimum wages are less than half the average wage (Fields (1994)). In Mexico, a substantial number of workers—even in large-scale enterprises—still earn less than the minimum wage (Fallon and Riveros (1989)).²⁴ A minimum wage law exists in Morocco but is not effective—in part because the high unemployment rate has led the authorities to allow employers some flexibility in hiring workers on a temporary and apprenticeship basis at wages below the minimum rate (Hollister and Goldstein (1994), p. 57). A study conducted in 1986 indicated that half the firms surveyed were paying their unskilled workers below the legal minimum (World Bank (1995). p. 75). In Tunisia, only 11 percent of the labor force is subject to the minimum wage (Hollister and Goldstein (1994), p. 60). Nevertheless, there is some evidence suggesting that the role of minimum wages can be significant. In countries like Sri Lanka, most private firms in the regulated sector pay wages near or above the minimum (Rama (1994a)). Camargo (1988) provides evidence suggesting that wages of unskilled workers in the formal sector are affected by the minimum wage, which he argued played a key role in determining a “wage floor” in the formal sector labor market in Brazil in the 1970s and early 1980s.²⁵ In more recent years, however, the minimum wage in Brazil has become increasingly less important in sectors other than services—particularly in manufacturing. Paldam and Riveros (1987) have used causality tests from minimum wages to average wages for a group of Latin American countries. Their results indicate that the causality is weak for Argentina and Penu, while it is stronger for Chile. Colombia, and Mexico. However, the results of such tests must be interpreted with care, since they are likely to be biased toward rejection of a significant link between minimum wages and average wages—unless minimum wages are adjusted frequently during the estimation period.

During the 1980s, minimum wages in most developing nations increased less rapidly than average wages or income per capita, and have declined in real terms in many countries—particularly in Africa and Latin America (Riveros (1990) and Fiszbein (1992)). In Mexico, the minimum wage fell by roughly 45 percent in real terms in the second half of the 1980s—while average real wages increased sharply in the aftermath of the implementation of the stabilization program in December 1987—whereas in Colombia the minimum wage fell by a much smaller amount between 1983 and 1987 (Bell (1994)). Real minimum wages in Uruguay fell by about 13 percent between 1981 and 1988. In Kenya, real minimum wages fell by more than 40 percent (and the average wage by almost 23 percent) between 1980 and 1986 (World Bank (1995) p. 76).

In many countries, the decline in real minimum wages has been large enough to erode the distortions their excessively high levels initially induced. However, despite this decline, the ratio of the minimum wage to unskilled labor wages—a more relevant indicator of the effect of minimum wages on the labor market—has not fallen significantly in a number of cases. Bell (1994), for instance, estimated the impact of minimum wages on the demand for skilled and unskilled labor in the formal manufacturing sector in Mexico and Colombia. At the end of the 1980s, the minimum wage stood at just 31 percent of the average unskilled manufacturing wage in Mexico, and roughly 53 percent of the average unskilled wage in Colombia. Bell found substantial disemployment effects of minimum wages in Colombia, with substantially larger effects for unskilled labor as opposed to skilled labor. She attributed the lack of evidence on disemployment effects in the case of Mexico to the relationship between the legally imposed minimum wage and the distribution of average unskilled wages across firms. She found that the minimum wage is very far to the left in the Mexico distribution and much closer to the mean in the Colombia distribution. Thus, minimum wages appeared ineffective in the formal manufacturing sector in Mexico and effective in Colombia.

Private and Public Sector Wages

Evidence on the behavior of urban, private sector wages shows marked differences across regions in the developing world. In many countries of Africa, real wages either stagnated or fell substantially during the 1980s. In Asia, sustained economic growth has often been associated with rising real wages, whereas in Latin America real wage behavior has shown considerable differences across countries.³⁶ In general, however, assessing the overall behavior of private sector earnings is hampered by the lack of data on wages in the informal sector. The available evidence for Latin America, for instance, suggests that although real wages in the modern sector grew in several countries during the 1980s (associated with a sharp reduction in inflation), the real income of those working in the informal sector fell, particularly among the self-employed. Ascertaining the determinants of the formal-informal wage ratio is thus a critical step in explaining the functioning of the labor market in developing countries. Some of the mechanisms through which private sector wages are formed and some of the available evidence pertaining to wage rigidity will be examined in the next section.

A number of recent studies have provided evidence on the evolution of public sector wages during the 1970s and 1980s. Kraay and Van Rijckeghem (1995) have examined data for a group of 27 developing countries over the past two decades. Their results indicate that central government wages represent about 4.8 times per capita income for the whole group of countries, and almost 5.6 for sub-Saharan Africa. Average wages in central government over the past two decades have provided a premium of about 20 percent relative to civil servants at the local government level, and about 10 percent relative to employees in public enterprises and the private sector. However, these data should be interpreted with care, because they are not adjusted for skills and other worker characteristics or for nonwage benefits.

Table 1, which is adapted from Hewitt and Van Rijckeghem (1995), presents data on central government wages based on a group of 99 countries over the period 1980–90. Wage expenditure of the central government in developing countries averaged about 7.5 percent of GDP (9.5 percent in the Middle East and North Africa, and 6.5 percent in Western Hemisphere countries), as opposed to 5.5 percent for industrial countries.³⁷ As a share of central government expenditures, wages averaged 27 percent for developing countries (28 percent for sub-Saharan Africa and 30 percent for Western Hemisphere countries), and 15 percent for industrial countries. The econometric estimates obtained by Hewitt and Van Rijckeghem suggest that these differences are related to the higher degree of decentralization in government spending in industrial countries. For the few countries for which data were available, general government wage expenditures amounted to 6.8 percent of GDP on average during the 1980s for developing countries and 11 percent for industrial countries. In percent of general government expenditure, wages and salaries amounted to 21.6 for developing countries and 22.6 for industrial countries. Industrial countries seem therefore to spend more on general government wage expenditures, although they spend less on central government wages.

A well-documented feature of the labor market in many developing countries during the late 1970s and early 1980s is the drop in real public sector salaries, and the relative stability of public employment. This reflects the fact that public sector expenditure cuts (which have often formed part of stabilization and adjustment programs) have tended to rely on pay, as opposed to personnel reductions in many of these countries. For the 15 largest Latin American countries (accounting for 95 percent of the region’s GDP), unweighted data show that between 1979 and 1988 public sector wages and salaries paid by the central government fell by about 0.4 percent of GDP (Hicks (1992)). But it is in Africa that the fall in public sector wages was the most dramatic. Lindauer, Meesook, and Suebsaeng (1988) have estimated that real wages for African government workers fell from the mid– 1970s through 1983 at a rate that often exceeded declines in per capita income.³⁸ Evidence of a significant erosion of public sector pay also exists for other countries, such as Egypt and Jordan (Hollister and Goldstein (1994), p. 64).³⁹ Such cuts have often had an adverse effect on morale and productivity, and have encouraged moonlighting (van Ginneken (1990) Stevenson (1992)). Compression of wage differentials in the public sector has also led to a shortage of skilled personnel in high-level positions, as well as to reduced incentives for employees in low-level positions to enhance their skills.

^{Table 1.}

Central Government Wage Erpenditure, 1980–90

(unweighted averages)

Source: Hewitt and Van Rijckeghem (1995), pp. 4 and 6.

^aTransition economies in Europe.

^{Table 1.}

Central Government Wage Erpenditure, 1980–90

(unweighted averages)

		1980–85	1985–90	1980–90
(in percent of GDP)
Total		7.3	7.1	7.2
Industrial countries		5.7	5.4	5.6
Developing countries		7.8	7.7	7.7
	Far East and Central Asia	6.2	6.9	6.5
	Middle East and North Africa	8.9	10.2	9.5
	Western Hemisphere	6.8	6.1	6.5
	Other a	6.7	7.0	6.8
Small low-income countries		7.4	6.7	7.1
(in percent of total central government expenditure)
Total		24.3	24.1	24.2
Industrial countries		15.5	15.1	15.3
Developing countries		27.0	26.8	26.9
	Far East and Central Asia	22.1	23.6	22.8
	Middle East and North Africa	22.8	28.1	25.2
	Western Hemisphere	30.2	29.6	30.0
	Other a	20.0	20.7	20.3
Small low-income countries		27.8	23.7	25.8

Source: Hewitt and Van Rijckeghem (1995), pp. 4 and 6.

^aTransition economies in Europe.

View Table

^{Table 1.}

Central Government Wage Erpenditure, 1980–90

(unweighted averages)

		1980–85	1985–90	1980–90
(in percent of GDP)
Total		7.3	7.1	7.2
Industrial countries		5.7	5.4	5.6
Developing countries		7.8	7.7	7.7
	Far East and Central Asia	6.2	6.9	6.5
	Middle East and North Africa	8.9	10.2	9.5
	Western Hemisphere	6.8	6.1	6.5
	Other a	6.7	7.0	6.8
Small low-income countries		7.4	6.7	7.1
(in percent of total central government expenditure)
Total		24.3	24.1	24.2
Industrial countries		15.5	15.1	15.3
Developing countries		27.0	26.8	26.9
	Far East and Central Asia	22.1	23.6	22.8
	Middle East and North Africa	22.8	28.1	25.2
	Western Hemisphere	30.2	29.6	30.0
	Other a	20.0	20.7	20.3
Small low-income countries		27.8	23.7	25.8

Source: Hewitt and Van Rijckeghem (1995), pp. 4 and 6.

^aTransition economies in Europe.

More recent evidence on the behavior of real wages in the public sector in Africa suggests that the dramatic fall that occurred in the early 1980s was subsequently reversed. The data compiled by Sahn (1992) suggest that although real wages in sub-Saharan Africa fell during 1984–86 as compared with 1980–83, they increased sharply during 1987–89—with the exception of non-oil exporting and CFA countries. In the CFA countries, there was in fact no reduction in real wages during 1984–86, and during 1987–89 real wages remained at the same level as in 1980–83. In proportion of government expenditure (net of interest payments), public sector wages and salaries dropped slightly between 1980–83 and 1984–86, but then increased significantly (to 36.7 percent) in CFA countries—thus crowding out other types of spending, particularly other recurrent expenditures (such as spending on materials or operations and maintenance). By contrast, in non-CFA countries, the share of wages and salaries fell continuously during the 1980s. The evidence presented by Sahn (1992) is consistent with the data gathered by Nashashibi and Bazzoni (1994), which indicate that public sector wage expenditures in fixed-exchange-rate sub-Saharan countries fell slightly in the first part of the 1980s (from 7.8 percent of GDP in 1980–81 to 7.6 percent in 1985–86) but increased thereafter (to 8.7 percent in 1990–91). By contrast, in variable-rate countries, wages fell from 6.9 percent of GDP in 1980–81 to 6.1 percent in 1985–86 and 5.8 percent in 1990–91.⁴⁰

Government pay and employment policies affect private labor markets through a variety of channels. In many countries, public sector wages and salaries in administration are low in comparison with private sector salaries—which makes it difficult to attract and retain qualified workers.⁴¹ In the first part of the 1980s, public sector workers in Latin America appear to have been somewhat less well paid than their private sector counterparts (López and Riveros (1989)).⁴² Public-private wage differentials appear to be much higher in Asia and Africa. Van der Gaag and Vijverberg (1988), for instance, estimate that public sector wages were much below private wages in Côte d’ Ivoire in the mid–1980s. In some other countries, however, public sector wages have been above those in the private sector.⁴³ In Sri Lanka, wages in the public sector are substantially higher than in the private sector (Dickens and Lang (1995)).

The data compiled by Kraay and Van Rijckeghem (1995) for a large sample of developing countries suggest that the premium of central government wages over private sector wages is about 9 percent, with sharp differences across countries. The ratio of central government wages to private wages is 1.22 in India and 1.38 in Botswana, compared with 0.6 in Ghana. Wage scales in the parastatal sector alone (essentially, financial and nonfinancial public enterprises) may be as much as 50 percent higher than in the private sector (Buffie (1994)).⁴⁴ There also appears to be little correlation between the rates of growth of real wages in the public and private sectors. Since the early 1980s the differential between public and private sector wages (in the formal sector) has either narrowed, as for instance in Kenya, or has continued to increase, as in Egypt (Horton, Kanbur, and Mazumdar (1994)). In Egypt, real wages during the 1980s fell more in the public sector than in the private sector, and within the public sector they fell faster for government workers than for public enterprise workers (Hollister and Goldstein (1994), p. 63).

Public sector employment may have a limited effect on market wages when labor is hired at wages below market rates (possibly in exchange for job security). The econometric results reported by Rama (1994a) for Sri Lanka, for instance, did not provide much evidence of a “wage leadership effect” by the public sector. Wage increases in both the regulated manufacturing sector and in nonregulated activities did not appear to be significantly correlated with wage movements in the public sector.

Nominal and Real Wage Inertia

Factors commonly viewed as accounting for nominal and real wage inertia in developing countries include minimum wages, indexation laws, employment protection provisions (such as labor tenure laws), restrictions on labor mobility (physical or otherwise), and large and powerful trade unions. Although the relative importance of these factors varies considerably across countries and over time,⁴⁶ an endemic feature has been implicit or explicit wage indexation. In high inflation countries in particular, wage indexation is an essential feature of the labor market.

However, despite the existence of various inhibiting factors, real wages in the developing world seem to be far more flexible than generally assumed. Horton, Kanbur, and Mazumdar (1994), in particular, summarize the findings of a comprehensive World Bank study on labor markets and adjustment in developing countries that provides evidence supporting the existence of a relatively high degree of real wage flexibility in Latin America and Asia. Evidence of downward real wage flexibility is also provided by Fallon and Riveros (1989) and by Hollister and Goldstein (1994) for countries of the Near East.

With flexible real wages, what explains the emergence and persistence of unemployment’? One line of reasoning (as discussed by Horton, Kanbur, and Mazumdar (1994)) is to attribute the persistence of unemployment to aggregate demand effects resulting from declining real wages and output market imperfections. The first type of effect has been emphasized by New Structuralist economists and is known as the Keynes-Kalecki effect.⁴⁷ The second type of effect may occur as a result of imperfect competition in product markets, even if labor markets are competitive and real wages flexible (Layard, Nickell, and Jackman 1991)). Another line of reasoning, as discussed below, is to rely on a combination of “wait” considerations, partial wage flexibility, and relative wage rigidity across skill categories.

Rather than absolute real wage rigidity, the labor market in developing countries may be characterized by nominal wage rigidity and/or relative wage inertia. Nominal wage rigidity results from a variety of factors, including staggered and overlapping wage contracts, and slow adjustment in inflationary expectations. The existence of multiperiod labor contracts is a particularly important source of nominal wage rigidity in some developing countries, particularly in Latin America. For instance, the evidence provided by Reinhart and Reinhart (1991) for Colombia (where the average length of labor contracts in the private sector is estimated to be about two years) suggests that there exists considerable inertia in the behavior of nominal wages. Studies by Allen, Cassoni, and Labadie (1994) on Chile and Uruguay; Condon, Corbo, and de Melo (1990), Corbo (1985), and Le Fort (1988) on Chile; and Kim (1994), Mazumdar (1993), and Van Wijnbergen (1985) 011 Korea also support the view that nominal wages are sticky in the short run. Most of these studies are based on the estimation of aggregate standard or extended Phillips curve equations, which relate the rate of growth of nominal wages to unemployment or output deviations from trend (a measure of the excess demand for labor), the expected inflation rate, and variables such as the lagged real wage. ⁴⁸ In the long run, standard specifications of Phillips curve relationships suggest that there should be no long-run tradeoff between inflation and unemployment—a view that appears to be supported by the data displayed in Figure 3. The coefficient on the unemployment or output gap variable gives an indication of the degree of wage rigidity (Hell iwell (1988)).⁴⁹ A problem with most of the existing literature on wage equations in developing countries, however, is that very few studies properly account for the time-series properties of the variables used in estimation, as emphasized in recent developments on cointegration.

Relative wage rigidity (across sectors and/or across skill categories), which is often a manifestation of labor market segmentation, may result from a variety of factors. With heterogeneous production structures across sectors, wage inertia may be the result of considerations internal to firms. In the presence of efficiency considerations (related to the existence of a wage-productivity link, shirking, or the existence of large turnover costs, as indicated earlier) firms may find it optimal to maintain a positive differential between wages offered to some categories of workers (typically, skilled workers, whose level of effort is often difficult to observe), and the going wage in alternative opportunities. In the framework developed by Agénor and Aizenman (1994), firms in the formal sector must pay skilled workers a fixed markup over the going wage in the informal sector, in order to induce the required level of effort.⁵⁰ In the model developed by Agénor and Aizenman (1996b), the higher wage earned by skilled workers in the formal or “primary” sector (relative to the informal or “secondary” sector wage) is also justified by the need to reduce incentives to shirk. In Agénor and Aizenman (1996a), the positive relation between formal and informal sector wages is related to the need to reduce turnover costs. Relative wage rigidity may also be related to the behavior of trade unions: in the macroeconomic framework developed in Agénor (1995c), the union’s target wage depends on the informal sector wage: as a result, the wage paid to skilled workers is positively correlated with wages in the informal labor market.

Inflation and Unemployment in Developing Countries

(Period averages, in percent)

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Sources: ILO, Yearbook of Labor Statistics; and IMF, International Financial Statistics.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Inflation and Unemployment in Developing Countries

(Period averages, in percent)

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Sources: ILO, Yearbook of Labor Statistics; and IMF, International Financial Statistics.

• Download Figure
• Download figure as PowerPoint slide

Inflation and Unemployment in Developing Countries

(Period averages, in percent)

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Sources: ILO, Yearbook of Labor Statistics; and IMF, International Financial Statistics.

• Download Figure
• Download figure as PowerPoint slide

Real Wages in Disinflation Programs

Various forms of wage policy have been used in stabilization programs implemented in developing countries. Argentina’s Austral Plan of June 1985 was preceded by a 22 percent increase in wages and a subsequent freeze. When prices kept rising—although at a much slower pace—the authorities allowed nominal wages to rise by 8.5 percent by the end of the year and then adopted quarterly wage adjustments. Israel’s stabilization plan of July 1985 granted a 50 percent compensation for that month’s inflation, and then froze wages for three months in a trilateral agreement between the Government, the entrepreneurs’ association, and the workers’ federation, the Histadrut (Arstein and Sussman (1990)). Subsequent adjustments provided partial compensation for previous inflation of 4 percent or more. At the inception of Bolivia’s August 1985 plan, the Government granted bonuses and then froze wages. Later, it reduced restrictions on laying off workers, eliminated wage indexation, and set a very low minimum wage. Brazil’s Cruzado Plan of February 1986 established an initial bonus of 8 percent of wages for all workers. At the same time, the minimum wage was increased by 16 percent. Nominal wages were not frozen, and annual (instead of semiannual) wage negotiations were restored. Wages were to be automatically adjusted when inflation reached 20 percent. This trigger was activated for the first time in December 1986, when the Cruzado Plan collapsed—in part as a result of excessive increases in public sector wages. As in Israel, Mexico’s stabilization program implemented in end–1987/early 1988 also relied on a collective agreement (the Pacto) between labor, employers, and the government. Thus, while in some cases initial wage increases were followed by a unilateral wage freeze in the public sector (eventually followed by further adjustments, such as in Bolivia), in other cases wage fixing and adjustments were based on a more or less implicit social contract between workers and the state (the cases of Argentina and Brazil) or an explicit agreement between workers, entrepreneurs, and the government, as in Israel and Mexico.

The behavior of real wages in disinflation programs in developing countries has received surprisingly little attention. Studies of exchange-rate-based programs have tended to focus on explaining the behavior of output on the basis of intertemporal effects associated with consumption behavior, with little emphasis on the supply-side effects of stabilization.⁵¹ However, recent analytical developments have emphasized that the long–term effects of stabilization policy may depend crucially on the nature of wage contracts. A reduction in the rate of nominal devaluation may lead in the long run to a contraction in output of tradables with backward-looking nominal wage contracts, but to an expansion in activity with forward-looking contracts (Agénor (1995a)).

The short-run dynamics of real wages in an exchange-rate-based stabilization program also depends crucially on the nature of wage contracts. If nominal wage contracts are backward-looking, a reduction in price inflation would lead at first to an increase in the real wage followed by a gradual reduction over time, as contracts begin to reflect the lower path of the inflation rate. However, the initial increase in the real wage may exacerbate the costs associated with stabilization.⁵² Indeed, a critical lesson from the experience of several Latin American countries in the early 1980s (Chile in particular) is that stabilization programs combining a fixed nominal exchange rate with backward-looking wage indexation leads to inflation inertia and results in an accelerating real appreciation of the exchange rate and an unsustainable widening of the current account deficit—often culminating in a balance of payments crisis and an exchange rate collapse.

If wage contracts are forward-looking, an anticipated future reduction in inflation that carries full credibility may lead to an immediate fall in the real wage if nominal wages are adjusted immediately to reflect the lower future path of prices. By contrast, if price and wage setters do not believe that the future reduction in prices will take place—for instance because the announcement of the future policy shift is not credible, or because agents expect the initial disinflation measures to be reversed in the future—nominal wages will not adjust, and the real wage may show little response. In fact, if future economic conditions are expected to deteriorate the real wage may rise immediately—despite initial corrective measures.

The Appendix presents a simple analytical model with backward- and forward-looking contracts that illustrates the above ideas. The model is based on a price equation relating changes in the inflation rate π_t, to excess demand for goods on the one hand, and a wage equation relating nominal wages to past values of prices (under backward-looking contracts) or future values of the price level (under forward-looking contracts) on the other. Simple manipulations lead to a dynamic system in terms of the inflation rate and the level of real wages, ω_t. The steady-state solution is such that the inflation rate and the real wage are constant and excess demand is zero. Regardless of the type of wage contracts, both the inflation rate and the rate of growth of nominal wages are equal to the devaluation rate ϵ^h in the long run.

Figure 4 shows the dynamic behavior of the model under backward- and forward-looking contracts with II (WW)representing the combinations of the inflation rate and the real wage for which the inflation rate (the real wage) does not change over time. The saddlepath is denoted SS and has a positive slope. In both cases, the steady-state solution obtains at point E.

Consider now the effect of a disinflation program that takes the form of a permanent, unanticipated reduction in the devaluation rate, from ϵ^h to ${ϵ}^s<{ϵ}^h$. The dynamics of wages and inflation are shown in Figure 5. In both cases the curve II shifts to the left (to I^’I^’) and the inflation rate falls continuously during the adjustment process. With backward-looking contracts, neither inflation nor the real wage changes on impact. Over time, inflation falls gradually but this tends to push the real wage up. The real wage increases monotonically toward its higher steady-state level, which is reached at point E^’. By contrast, with forward-looking contracts, and since agents discount the future reduction in inflation back to the present, the real wage jumps downward immediately to a point such as A on the new saddlepath S^’ S^’ shown in the lower panel of Figure 5 and continues to fall toward its lower steady-state level, which is also reached at E^’. Thus, while the adjustment process to a cut in the devaluation rate leads to a gradual increase in real wages with backward-looking contracts, it leads to an initial downward jump followed by a continuous fall in real wages with forward- looking contracts.

Steady-State Equilibium with Alternative Wage Contracts

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Steady-State Equilibium with Alternative Wage Contracts

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

Steady-State Equilibium with Alternative Wage Contracts

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

Disinflation and Real Wage Dynamics

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Disinflation and Real Wage Dynamics

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

Disinflation and Real Wage Dynamics

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

What does the evidence suggest regarding the evolution of real wages in the type of exchange-rate based stabilization programs analyzed above? Figures 6 and 7, which relate to the Tablita experiments of the late 1970s in Latin America and the “heterodox” experiments of the mid–1980s in Israel and Mexico, provide a mixed picture in that regard. At the inception of the Tablita experiments, real wages remained either stable (Uruguay) or rose, whereas they fell in the orthodox experiments.⁵³ The extent to which such movements can be viewed as reflecting backward- or forward-looking behavior and/or lack of credibility cannot be assessed without appropriate econometric methods. Unfortunately, despite the importance of this issue for stabilization policy, there have been very few attempts at estimating the degree to which wage formation is backward- or forward-looking in developing countries. Extensions of the methodology proposed by Moghadam and Wren-Lewis (1994) might prove useful in that regard.

Stabilization with Segmented Labor Markets: A Formal Framework

The early emphasis on the role of real wage resistance in determining the short- and longer-run effects of stabilization and adjustment policies has given ground in recent years to an emphasis on the role of labor market segmentation (and the type of rigidities that it creates) in determining the response of wages and employment to macroeconomic policy in developing countries.

Segmentation often involves different degrees of wage rigidity across alternative segments of the labor market and labor categories. In an early study on this issue, López and Riveros (1989) have argued that labor market segmentation alters the differential between formal and informal wages during adjustment, and may render a nominal devaluation ineffective in altering the real exchange rate. In their model, wage indexation prevails in the formal (tradable) sector, whereas wages in the informal (nontradable) sector are fully flexible. Following a nominal devaluation, workers in the formal sector will thus be able to maintain their real wages relative to workers in the informal sector. In such conditions, the devaluation will mainly lead to a widening of the Formal-informal wage differential and a smaller real depreciation.⁵⁴ In their review of the labor market in Argentina. Colombia, Chile, and Uruguay, López and Riveros (1989) argue that formal sector wages are indeed more responsive to exchange rate changes than informal wages, and that an increase in the price of tradables implies a decline in real informal sector wages. Thus, a devaluation is likely to increase the formal-informal sector wage gap. The more rigid formal sector wages are, the larger the required nominal devaluation to achieve a given rate of depreciation of the real exchange rate.

Inflation and Real Wages in the Tablita Experiments

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Source: Végh (1992).Note: Shaded areas indicate periods during which the programs were in place.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Inflation and Real Wages in the Tablita Experiments

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Source: Végh (1992).Note: Shaded areas indicate periods during which the programs were in place.

• Download Figure
• Download figure as PowerPoint slide

Inflation and Real Wages in the Tablita Experiments

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Source: Végh (1992).Note: Shaded areas indicate periods during which the programs were in place.

• Download Figure
• Download figure as PowerPoint slide

Inflation and Real Wages in Heterodox Experiments

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Source: IMF, International Financial Statistics.Note: The vertical line indicates the start of the stabilization program.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Inflation and Real Wages in Heterodox Experiments

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Source: IMF, International Financial Statistics.Note: The vertical line indicates the start of the stabilization program.

• Download Figure
• Download figure as PowerPoint slide

Inflation and Real Wages in Heterodox Experiments

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

Source: IMF, International Financial Statistics.Note: The vertical line indicates the start of the stabilization program.

• Download Figure
• Download figure as PowerPoint slide

An important limitation of some recent macroeconomic models with segmented labor markets—including the framework developed by López and Riveros (1989)—is the absence of a fully articulated theory of the channels through which wages in the formal and informal sectors interact. While more recent papers have retained the assumption that a high degree of wage flexibility prevails in the informal sector, they have emphasized the role of efficiency considerations and explored in a more detailed manner the role of trade unions in explaining wage formation in the formal sector. The role of trade unions in wage formation in the formal sector is a central feature of the model developed by Agénor (1995c). In Agénor and Aizenman (1994 and 1996b), firms in the formal or “primary” segment of the labor market set wages above the level consistent with market clearing in order to elicit the required level of productivity from skilled workers and deter shirking—since monitoring on-the-job effort for that category of workers is often difficult. In Agénor and Aizenman (1996a), efficiency considerations by firms in the formal (tradable) sector result from the existence of high turnover costs, and a link between relative wages and the quit rate. Indeed, a growing literature supports the existence of efficiency considerations in the modern sector in developing countries.⁵⁵

The framework developed by Agénor (1995b) captures many of the economic, institutional, and regulatory features of labor markets in developing countries reviewed in Section I, and is convenient for analyzing the wage and employment effects of various stabilization and adjustment policies. Three categories of agents are assumed to operate in the economy: firms, households, and the government. The economy consists of two major segments: the formal economy and the informal sector. In the formal economy two goods are produced: an exportable good, whose output is entirely sold abroad, and a nontraded government service. Firms in the informal economy produce a nontraded good that is used only for final domestic consumption. The labor force is constant and consists of skilled and unskilled workers. Production of the nontraded good and government services requires only unskilled labor. To produce the export good, government services are used as intermediate inputs, together with both categories of labor.

The labor market is segmented as a result of government regulations and relative wage rigidity. Public sector employment and the real wage (measured in terms of exportables) earned by government employees are exogenously determined. In the export sector, the wage earned by skilled workers is positively related to the informal sector wage (as a result of efficiency considerations, or the existence of a utility-maximizing trade union), while unskilled workers earn the minimum wage fixed by the government. For a given level of wages, firms in that sector determine the level of employment of both categories of labor so as to maximize profits. Mobility of the unskilled labor force between the formal and the informal sectors is imperfect.⁵⁶ Migration flows are determined by expected income opportunities, along the lines of the Harris-Todaro model discussed earlier. Skilled workers who are not able to get a job in the formal sector are not necessarily willing to take employment (as unskilled workers) in the informal economy. Whether or not they do depends on the perceived net cost of working there (related to factors such as demotivation, the efficiency of on- the-job searches, loss of skills or social prestige, and the extent to which relatives are capable of providing a “safety net”). In the informal sector, which absorbs all unskilled workers who do not queue up for employment in the formal sector as well as skilled workers who are willing to work there, wages adjust continuously to equilibrate supply and demand for labor. In both the short and the long run, therefore, “luxury” unemployment of skilled workers and “wait” unemployment of unskilled workers may emerge.⁵⁷ Prices are also flexible in the informal sector, and adjust to eliminate excess demand for nontraded goods.

Households supply labor inelastically and consume, in addition to the nontraded good produced in the informal sector, an imported good that is imperfectly substitutable for the home good. Total consumption expenditure is linear in financial wealth, which consists of traded bonds accumulated through current account surpluses. Wage and employment prospects are formed on the basis of prevailing conditions in the labor market.

As indicated earlier, the government produces services that are used as intermediate inputs in the production of export goods. It sets the real price (measured in terms of export goods) of its output, hires a constant share of the unskilled labor force, and fixes production at a level that is more than sufficient to satisfy private demand for government services—thereby creating “labor hoarding” (or disguised unemployment) in the public sector. It also consumes imported goods, and it finances spending by levying lump- sum taxes on households.

The world price of exports is exogenous, and the production function in the export sector is characterized by fixed proportions between government services and gross output, and between a composite bundle of skilled and unskilled labor, which produces value added. Thus, there are no substitution possibilities between government services, on the one hand, and the composite bundle of labor services, on the other. Skilled and unskilled labor are taken to be (Edgeworth) substitutes in the production of value added.⁵⁸ Under the assumption of profit maximization, the demand functions for labor and government services are such that an increase in the real wage earned by skilled workers reduces the demand for that category of labor and raises the demand for unskilled labor, whereas an increase in the minimum wage has the opposite effects. A rise in the real price of government services reduces the demand for both categories of labor.

Technology for the production of the nontraded good in the informal sector is characterized by decreasing returns to labor. Profit maximization yields the familiar equality between marginal revenue and marginal cost, from which labor demand is derived. The supply function for goods produced in the informal sector is then negatively related to the product wage, defined as the real informal sector wage (measured in terms of the price of exports) times the real exchange rate—itself defined as the price of exports relative to the price of nontraded goods.

The process of wage formation varies across segments of the labor market. As indicated above, both the minimum wage paid to unskilled workers in the export sector and public sector salaries are fixed by the government in real terms. The wage earned by skilled workers in the export sector is positively related to the informal sector wage. As discussed below, the elasticity of the skilled worker wage to the informal sector wage (denoted as γ in what follows) plays a critical role in the dynamics of policy shocks.

Consider now the informal sector labor market. The demand for labor is derived from profit maximization. The supply of unskilled workers in the formal sector is predetermined at any moment in time, and changes gradually as a response to the expected wage differential across sectors⁵⁹ Assuming for simplicity that government workers are paid the minimum wage, the expected wage in the formal economy is equal to the minimum wage times the probability of being hired. Assuming that hiring is random, the hiring probability can be approximated by the prevailing employment ratio. The expected wage in the informal economy is simply the going wage, since there are no barriers to entry in that sector. Skilled workers queue up for employment in the formal sector first, but those who are unable to find a job in that sector may or may not seek employment in non- traded goods activities. In line with the evidence reviewed in Section I, we focus here on the case where skilled workers who are unsuccessful in finding a job in the formal sector opt to remain unemployed. Thus, the informal sector wage adjusted for the disutility of effort—that is, the opportunity cost of “leisure”—is lower than the expected return from being unemployed. The equilibrium condition of the labor market in the informal economy implies that the informal sector real wage (measured in terms of the price of exports) is inversely related to the real exchange rate (the relative price of traded goods) and positively related to the size of the population of unskilled workers seeking employment in the formal economy.

The dynamics of the model can be formulated in terms of the size of the unskilled labor force seeking employment in the formal economy, and households’ holdings of traded bonds expressed in foreign currency terms. The “short-run” variables, the real exchange rate and the real informal sector wage, are solved jointly through the equilibrium conditions of the market for nontraded goods and the informal labor market. In the steady-state equilibrium, the ratio of wages earned by unskilled workers in the formal and informal sectors (the unskilled wage ratio, for short) is equal to the inverse of the employment ratio of that category of labor in the formal economy.

A simple graphical illustration of labor market equilibrium in the above model is presented in Figures 8 and 9. Figure 8 shows the case in which unskilled labor is perfectly mobile across sectors, whereas Figure 9 corresponds to the case in which the size of the unskilled labor force seeking employment in the formal economy is determined through a Harris-Todaro mechanism. In both figures, panel A presents labor market equilibrium in the formal sector. The demand for skilled labor $L_{FS}^d$ is inversely related to $w_F^S$, the wage earned by skilled workers. The total demand for unskilled labor is $L_{FU}^d$, which is equal to the sum of the demand for unskilled workers by private firms in the export sector (which is positively related to $w_F^S$, since skilled and unskilled workers are gross substitutes) and public sector employment (which is exogenous). By subtracting $L_{FU}^d$from the total supply of unskilled workers ${\overline{L}}_U$, the supply of labor in the informal economy is determined in panel $B$. Given the labor demand curve in the informal sector $L_I^d$, the market-clearing wage is determined at point C in panel C. The positive relationship between the skilled workers’ wage and the informal sector wage is displayed as curve WW in panel D. As shown in panel A, unemployment of skilled workers prevails in equilibrium (since unsuccessful applicants prefer not to work in the informal economy) and is given by the distance between the total supply of skilled labor ${\overline{L}}_S$ and the equilibrium point on the demand curve.

Figure 9 illustrates the case in which unemployment of both categories of labor prevails in equilibrium. The key difference with the previous case is that, as discussed earlier, in a Harris-Todaro setting the supply of unskilled workers in the formal sector, $L_{FU}^s$, is proportional to the demand for labor in that sector times the unskilled wage ratio in equilibrium. If the unskilled wage ratio is greater than unity, that is, if the informal sector wage is lower than the minimum wage—a necessary condition to ensure that unskilled workers will queue for formal sector jobs in the first place—$L_{FU}^s$ will be greater than $L_{FU}^d$ and unemployment of unskilled labor will also prevail in equilibrium, as shown in panel B.

Labor Market Equilibrium with “LUXURY” Unemployment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Labor Market Equilibrium with “LUXURY” Unemployment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

Labor Market Equilibrium with “LUXURY” Unemployment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

Labor Market Equilibrium with Generalized Unemployment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Labor Market Equilibrium with Generalized Unemployment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

Labor Market Equilibrium with Generalized Unemployment

Citation: IMF Staff Papers 1996, 002; 10.5089/9781451947229.024.A001

• Download Figure
• Download figure as PowerPoint slide

While the model described above can be used to analyze a large variety of policy shocks (see Agé(1995c)), we will discuss here only the wage and employment effects of public sector layoffs. As indicated in Section I, policies involving the retrenchment of public sector employment have indeed often been part of both stabilization and structural adjustment programs, but very few studies have attempted to study their macroeconomic effects in a framework that accounts for labor market segmentation. In the above setup, a reduction in public sector employment leads in the long run to a reduction in the number of unskilled job seekers in the formal sector. The resulting increase in labor supply in the informal sector lowers wages in that sector and raises output of nontraded goods. Skilled employment and output in the export sector also rise (as long as $\gamma >0$). Skilled unemployment is lower, but because the total demand for unskilled workers falls by more than the reduction in labor supply in the formal sector, the unskilled unemployment rate is higher in the new steady state. On impact, the real exchange rate and the informal sector wage do not change. The reduction in public sector employment nevertheless lowers the expected wage in the formal sector and reduces incentives for unskilled workers to wait for jobs in the formal sector. Over time, the increase in labor supply lowers real wages in the informal sector—thereby raising output in that sector and leading to a depreciation of the real exchange rate—but the continuous reduction in the demand for unskilled labor in the formal economy (resulting, during the transition, from the substitution toward skilled labor in the export sector, as a result of the reduction in the skilled workers’ wage) tends to dominate, thereby leading to a continuous fall in the number of job seekers in the formal sector. Thus, public sector layoffs may lead in the long run to a switch toward informal activities.⁶⁰ Evidence showing that cuts in public sector employment (or reductions in recruitment rates) have led in some countries to higher employment in informal activities has indeed been documented in several recent studies, such as Hollister and Goldstein ((1994), p. 62).⁶¹

The analytical framework presented above helps explain the apparent weak relationship (discussed earlier) between unemployment and output growth in developing countries. Interactions between the formal and informal urban labor markets may be characterized by substitutability rather than complementarity in the short run, implying that the employment effects of macroeconomic policy shocks can be highly mitigated (Agénor and Aizenman (1994)). In periods of weak output growth, unskilled workers (and, to a lesser extent, skilled workers) laid off in the formal sector may seek employment in the informal sector where wages and labor productivity tend to be lower. Thus, output fluctuations may translate into changes in average productivity rather than increases in open unemployment.

Labor Market Reforms and Labor Market Flexibility

Advocates of labor market reform have argued that circumventing the scope of hiring and firing regulations, reducing nonwage labor costs, eliminating (or restricting the scope of) minimum wage laws, limiting unemployment benefits, and curtailing the role of trade unions in the wage bargaining process would enhance flexibility and have a positive effect on growth. However, a balanced view of the effects of such reforms must consider a variety of conflicting effects.

It is widely believed that job security provisions, despite often being commendable in principle (to the extent that their objective is the protection of workers against unsafe work practices and unjustified dismissals) have had in practice a variety of adverse consequences. Such provisions have led in some cases to a general loss of profitability, reduced flexibility at the firm level to relative price shocks and shifts in demand, and distortions in favor of more capital-intensive production techniques (Fallon and Riveros (1989)).⁶³ Enterprises tend to become reluctant to take on new employees when faced with high hiring and firing costs. In addition, employers may become cautious about hiring new workers with contracts of indefinite duration, opting instead to rely on casual labor, subcontracting, or fixed-term contractual relationships.⁶⁴ By inhibiting the reallocation of labor and wage flexibility, labor regulations lead to higher (and more persistent) unemployment.

The role of hiring and firing costs in introducing inertia into firms’ employment decisions has been emphasized also in the context of industrial countries, notably by Bean (1994) and Lazear (1990). Bean has argued that such costs (together with the progressive disconnection of the unemployed from the labor market) are the major factor explaining the persistence of unemployment in European countries. By increasing the incidence of long-term unemployment through reduced labor turnover, they speed up the “deskilling” process among the unemployed and reduce the downward pressure exerted by the unemployed on wages.

However, this “consensus” view has become the subject of intense controversy in recent years, in both industrial and developing countries. For industrial countries (particularly in Europe), there appears to be limited empirical evidence supporting the view that employment protection measures (such as hiring and dismissal procedures) are sources of unemployment persistence. A careful examination of employment protection measures in Germany, France, and Belgium in the mid–1980s by Abraham and Houseman (1994) found no evidence that changes in such provisions affected the speed of labor market adjustment.⁶⁵ Recent research on employers’ responses to the introduction of unfair dismissal legislation in the United Kingdom also showed little evidence of reduced hiring. In the context of developing countries, Freeman (1993) has argued that high hiring and firing costs may have more limited practical effects than is often thought where (at least segments of) labor markets are highly flexible—precisely in part due to poor compliance with existing regulations. Job security provisions may have no direct effects on employment and unemployment, if wages are de facto flexible. Indeed, it has been argued that standardization of the contractual rules across firms may reduce information and transactions costs, thus rendering the market more efficient. However, if employment protection legislation is not applied uniformly to all firms in all production sectors (or if enforcement varies according to firm size), it will distort labor allocation and foster the expansion of the informal economy, as producers go underground.

In recent years, Standing (1991) has been one of the most forceful advocates of the positive effects of labor market regulations. In particular, such regulations can improve workers’ commitment to the enterprise and thus raise work motivation and productivity. They may reduce the transactions costs of employment, by reducing labor turnover—a consideration that carries particular importance when productivity rises with on-the-job learning. They may improve job and work flexibility, that is, improve the willingness of workers to accept (and even initiate) occupational and work environment changes. They may induce workers to accept lower wage rises. Finally, they may reduce the probability of frictional unemployment by enabling workers made redundant to have adequate notice of impending job loss to seek alternative employment, thus reducing both the individual and social costs of mobility. Determining the net cost of employment security regulations requires taking into account all their potential beneflts.⁶⁶ In a careful review of the experience of several industrial countries, for instance. Blank and Freeman (1994) have found that there is little empirical evidence supporting the existence of a large trade-off between labor market flexibility and social protection programs. Job security provisions—advance notice and severance pay requirements—in Western Europe (Belgium, France, and Germany) have not hindered labor market flexibility.⁶⁷ Another recent collective study by the Center for Economic Policy Research suggests that dismantling labor regulations would be a naive response to high unemployment in Europe, as the cost of these restrictions is not as high as it looks (Alogoskouf is and others (1995)).

The role of minimum wage laws on the functioning of labor markets in developing countries continues to be the subject of intense controversy. Two divergent views appear to dominate the scene (Freeman (1992)). The “advocate” view holds that minimum wages have positive nutritional effects in poor countries—as emphasized in the efficiency-wage model of Dasgupta and Ray (1986), for instance—and may redistribute resources in a welfare-enhancing way. As a result, high minimum wages may help reduce poverty and raise productivity by enticing workers to work harder and change jobs less frequently.⁶⁸ They may also enhance growth prospects and increase welfare if the positive externality associated with human capital accumulation has a sufficiently large impact on overall productivity (Cahuc and Michel (1996)).

The “distortionist” view, by contrast, suggests that minimum wage legislation leads to excessively high wage costs (particularly for unskilled labor) in the formal sector, misallocation of labor and lower employment (by preventing wages from adjusting downward to excess supply of labor), the creation of rents for some groups of workers (Rama and Tabellini, 1995), labor market segmentation (as in the Harris-Todaro model presented earlier), depressed wages in the informal urban sector (which has an adverse effect on the poor), and ultimately a reduction in the rate of growth. Even though a higher minimum wage may raise employment in the short term, it also increases the relative cost of employing unskilled workers (thereby accelerating the substitution of capital for unskilled labor) and reduces profits—thereby lowering firms’ capacity to invest. As a result of both factors, high minimum wages restrain the expansion of labor demand over time.⁶⁹

In general, assessing the effect of minimum wages on average wages and employment of unskilled workers is a difficult task in developing countries. Lack of compliance with the law makes studies based on the formal sector highly imprecise. As indicated earlier, for instance, a significant share of workers in Mexico are paid below minimum wages in the formal manufacturing sector. Workers in some segments of the formal manufacturing sector may also earn wages in excess of the minimum wage. At least in theory, changes in minimum wages may have important distributional effects within the category of unskilled workers,’ for instance between those employed in the formal sector and the informal sector. Explicit consideration of the informal sector (as was done earlier) is thus an essential step in analyzing the aggregate effects of changes in the minimum wage.

It has been argued that unemployment insurance or compensation schemes may act as a disincentive to search for (or to accept) employment, and that they may act as an incentive to enter the labor force solely to collect unemployment benefits. However, unemployment insurance may also have positive effects in encouraging labor force participation and favoring regular, as opposed to marginal, employment (Atkinson and Micklewright (1991)). Because of limited data on variables such as the duration of unemployment benefits in developing countries, it has proved difficult to test alternative views on these issues. The elasticity of unemployment with respect to replacement rates (benefits before taxes as a percentage of previous earnings) may be relatively low, but whether high unemployment benefits tend to increase open unemployment remains an open issue.

In assessing the role of trade unions in the bargaining process and wage formation in developing countries, two considerations are important (Nelson (1994)). First, trade union movements in these countries are typically not very centralized (even in Latin America), making it difficult to organize collective labor action. In Mexico, for instance, the average union size in the Confederation of Mexican Workers (the country’s largest confederation) is small (about 130 members). Less than half of the Confederation’s members are in powerful industrial unions (such as railroads, petroleum, or mining). Nearly half of them are in enterprise-level unions, which are, in turn, loosely related to regional federations. However, the aggregate implications of a low degree of centralization on wage formation are not entirely clear. Economic theory suggests that the relationship between the degree of centralization in wage bargaining—defined as the extent to which unions and employers cooperate in wage negotiations—and wage pressures has an inverted U-shape, rather than being monotonic. Wage push is limited when bargaining is highly centralized (conducted at the union level) and highly decentralized—when it takes place at the level of individual firms (Calmfors (1993); Moene, Wallerstein, and Hoel (1993)). The highest degree of influence on wage formation would thus tend to occur in countries where centralization is in the intermediate range, that is, at the industry level.⁷⁰ In practice, however, it is not always easy to classify wage bargaining systems into completely centralized or decentralized systems. In Argentina in recent years, the Government has sought to decentralize collective bargaining agreements from the sectoral to the firm level. In early 1995 Mexico also moved toward terminating centralized wage agreements in favor of decentralized bargaining arrangements. In Brazil, by contrast, the evidence seems to suggest that wage negotiations have become more centralized in a number of industries during the 1980s (Amadeo (1994)). Whether the outcome in these countries will be a reduction in wage pressures is unclear at this stage.

The second consideration is that the degree of unionization is a highly imperfect measure of the influence of trade unions on wage formation and the labor market. Trade unions in certain “strategic” sectors or industries may exert considerable influence on wage formation and working conditions at the national level, even if overall union membership is a low proportion of the workforce. Often-cited examples are the trade unions representing tin miners in Bolivia and oil workers in Nigeria (Nelson (1994)). In Singapore, although union density is low, collective bargaining agreements are often extended to the nonunionized workers within individual firms (Carling (1995)).

The empirical evidence on the impact of unions on real wages in developing countries is somewhat limited. Nelson (1994) suggests that in Latin America unions have caused wages to rise above the opportunity cost of labor through a combination of union pressure, minimum wage legislation, and wage policies in the public sector. In Taiwan and Korea, the evidence suggests that unions have limited power in bargaining over wages (Fields (1994)). Other studies have attempted to estimate directly the bargaining strength of organized labor, as reflected in the union-nonunion wage differential. Park (1991), for instance, estimates that blue-collar workers in the unionized manufacturing sector in Korea are paid on average only 4 percent more than their counterparts in the nonunionized sector. Panagides and Patrinos (1994) have estimated that the union-nonunion wage differential in Mexico (based on a household survey for 1989) was about 10.4 percent, which indicates some bargaining strength.⁷¹ Amadeo (1994) provides evidence of a positive relationship between unions’ (relative) bargaining power—measured by the degree of indexation of wages to price changes—and relative wages in Brazil. More generally, however, the assumption that trade unions are often powerful enough to affect wage formation appears to be in contradiction with the fact that real wages fell quite significantly in a number of Latin American and sub-Saharan African countries during the 1980s, as discussed above. Thus, attempts at circumventing the role of unions in the process of wage formation are unlikely to have any significant effect on “flexibility” in most cases.

Trade Liberalization and Unemployment

In recent years, growing recognition of the severe allocative distortions associated with import substitution strategies has led an increasing number of countries in the developing world to adopt a more liberal external trade regime. Traditional economic arguments suggest that a reduction in trade barriers (such as tariffs and import quotas) fosters an adjustment in relative prices that leads to a reallocation of resources toward the exportable sector. In the long run, trade reform leads to an expansion of output of exportables and a contraction of activity in import-competing industries, as well as to an overall transfer of resources from sectors producing nontradables toward those producing tradables.

While there appears to be broad agreement on the allocative effects of trade liberalization in the long run, the short- and medium-run impact of trade reform on the wage structure, the composition of employment, and aggregate unemployment remains imperfectly understood. The evidence gathered in the comprehensive study of trade reform episodes in developing countries conducted in the early 1980s at the World Bank and summarized by Papageorgiou and others (1990) appears to be largely inconclusive in that regard. While most individual country studies suggested that total employment in the manufacturing sector either fell or remained stable in the aftermath of the liberalization program, they often did not distinguish between traded and nontraded manufacturing goods, and were therefore unable to characterize changes in the distribution of employment. In addition, almost no evidence was provided on changes in employment in production activities other than manufacturing, or changes in aggregate unemployment.⁷²

Sonic more recent studies have attempted to study directly the effects of tariff reform on the labor market. Rama (1994b) examined the relationship between tariffs, employment, and wages in the Uruguayan manufacturing sector. He found that the reform had no impact on wages, but a negative effect on employment. His estimates indicate that a reduction in the tariff-inclusive price of imports by 1 percentage point led to an employment drop in manufacturing of between 0.4 and 0.5 percentage points. In a study of the trade liberalization program implemented in Mexico between 1985 and 1988. Revenga (1994) estimated that the reduction in tariffs during the period (of about 10 percentage points) led to a much smaller reduction in aggregate employment in the manufacturing sector (by 2 to 3 percentage points) and an increase in average wages.⁷³ However, her study also suggested that, despite relatively limited aggregate effects, significant changes occurred in the composition of employment at the industry level. Currie and Harrison (1994) found that the comprehensive trade reform that was implemented in Morocco between 1984 and 1990 (which led to a reduction in the coverage of import licenses from 41 percent of imports in 1984 to 11 percent in 1990, and a reduction in the maximum tariff rate from 165 percent to 45 percent) also had a small (albeit significant) impact on aggregate wages and employment in the formal manufacturing sector. As in the case of Mexico, pronounced sectoral shifts in employment appeared to have taken place, particularly in the manufacturing industries that were subject to large tariff reductions.

Very few economists have attempted to examine in formal analytical models the impact of trade reform on the dynamics of wages, the composition of employment, and aggregate unemployment in the presence of labor market distortions. Among the few studies available are those of Buffie (1984), Edwards (1988), and Cox Edwards and Edwards (1994). However, these authors abstract from potentially important factors that may alter the impact of trade reform on the composition of employment and the aggregate unemployment rate in developing countries. Buffie (1984) focuses on the potential short-run contractionary effect of trade liberalization in the presence of economy-wide nominal and real wage rigidity, whereas Edwards (1988) considers in his analysis of tariff reform an economy in which wages are rigid in only one sector. Both authors, however, focus on the case of absolute, rather than relative, wage rigidity—thus excluding potential interactions between wage formation in different sectors of the economy—and do not account for impediments to labor mobility in the short run. The models considered by Cox Edwards and Edwards (1994) are also subject to the first type of limitations, although they do address the issue of intersectoral labor mobility. In addition, their treatment of the dynamics associated with trade liberalization does not allow a full characterization of the adjustment process.

In contrast to the previous literature, Agénor and Aizenman (1996a) account explicitly for imperfect labor mobility and interactions between wage formation across sectors, highlighting the role of efficiency considerations induced by the existence of high turnover costs in the formal (tradable) sector.⁷⁴ Their analysis suggests that a permanent reduction in tariffs (coupled with an adjustment in lump-sum taxes to equilibrate the government budget) may reduce unemployment in the short run, but may increase it in the long run. The model, however, is subject to some limitations and the robustness of some of the results must be assessed in more general settings. Specifically, it appears important to account for the existence of worker heterogeneity, along the lines discussed in Section II. In her study of trade liberalization in Mexico, for instance, Revenga (1994) noted that the observed increase in average manufacturing wages may have reflected a change in the composition of the labor force—a shift toward high-skill, high-wage workers. A similar shift in the composition of the workforce was noted by Currie and Harrison (1994) in their analysis of trade reform in Morocco. Understanding the mechanisms through which tariff reductions affect the skill composition of employment is particularly important for studying their distributional effects.⁷⁵