Privatization has been assessed in terms of increased efficiency versus increased unemployment. Proponents of privatization, such as the IMF and the World Bank, point out the presumed greater efficiency of private enterprises. Opponents of privatization urge caution because of the presumed greater unemployment that privatization would cause. The same argument buttresses the case for action and the case for caution; that public enterprises are inefficient in part because they use too much labor. In developing countries, the use of too much labor in production has often been called either surplus labor or disguised unemployment. The argument presumes first that too much labor is used in the public enterprises and that this surplus labor is the source of the inefficiency of public enterprises.

In this paper, we will examine the possibility that such a phenomenon exists in the publicly owned Egyptian textile industry. First, we will explain the notion of surplus labor and the history of this notion as it applies to Egypt. In the next section, an empirical model of surplus labor in government owned enterprises will be presented and tested. Finally, conclusions will be drawn concerning disguised unemployment and surplus labor in Egypt, and the policy implications of our findings for Egypt and elsewhere.

Surplus Labor and Disguised Unemployment

Surplus labor was thought largely to exist in the traditional, agricultural sector of less developed nations. The notions of disguised unemployment and surplus labor have been a central part of development economics since Lewis (1954). In dualistic models in this tradition, labor in agriculture is thought to earn its average product instead of the marginal product of labor. In many formulations it was assumed that in the agricultural sector MPL = 0 and, hence, that labor could be transferred from agriculture to manufacturing at no opportunity cost.

While the concepts of surplus labor and disguised unemployment are closely linked, we will distinguish between these two concepts for greater clarity and because surplus labor may not exist in Egypt in either the traditional or the government sector. Disguised unemployment will be said to exist if labor earns more than its marginal product, w > MPL. This would occur if too much labor was employed at the prevailing wage rate. Surplus labor will be said to exist if the marginal product of labor is less than the subsistence wage rate, s > MPL. Since by definition the wage rate must at least equal the subsistence wage rate, this implies that w > s > MPL. Hence, surplus labor implies disguised unemployment. However, disguised unemployment does not imply surplus labor because we could have w > MPL > s.

One explanation for why labor might be paid more than its marginal product relates to the nature of the family enterprises that are so prevalent in the traditional sector. Family workers share in the rents and profits of family enterprises but only if they work in the enterprise. If rents and profits are divided equally among family workers, each will earn the average product of labor, APL_t, where the subscript indicates the traditional sector. Consequently, workers will remain with the family enterprise as long as APL_t > w_m, where w_m = MPL_m is the wage rate in the modern sector. Since, presumably, w_m = MPL_m , MPL_m > MPL_t and too much labor is employed in the traditional sector.

Initially, in Egypt both surplus labor and, hence, disguised unemployment were thought to exist in agriculture. There are three reasons for believing that disguised unemployment existed in Egyptian agriculture. First, more than 70% of Egypt's population lived in rural farm land areas, that represented less than 1% of the total area of the country giving credence to the description of Egypt as a typical Malthusian economy where wages center around the subsistence level. Second, the farms were traditional, family farms. Consequently, they would employ more labor than the competitive neoclassical firm because in family firms the output of the firm, minus costs, is shared with family members. The share is based partly on the person's contribution to the production of the family enterprise. Consequently, the average family worker-owner earns the average product of labor. Second, the marginal product of labor in Egyptian agriculture was low. Indeed, using a weak disposability of inputs production function, Aly and Grabowski (1984) found a negative marginal product of labor in Egyptian agriculture from 1952 to 1972. Hence, the subsistence wage rate clearly exceeded the marginal product of labor because w > s > 0 >MPL.

Since the oil boom of 1973, this situation may have dramatically changed. The oil boom resulted in a sharp increase in the demand for workers throughout the Gulf region. As a result, there was a large migration of Egyptian workers to Gulf countries that dramatically reduced the supply of workers in Egypt. Richard and Waterbury (1990) estimate that 3.5 million Egyptians migrated during the period from 1973 to 1985. This migration has had a considerable impact on the Egyptian economy. One impact has been to increase living standards in agriculture. Indeed, the Egyptian economy has undergone a dramatic transformation. GDP grew by 9.3% a year from 1976 to 1980 and by 5.0% from, 1980-1990. The higher wage income resulting from this transformation puts the notion of surplus labor in agriculture in doubt.

Actually, this agricultural transformation has been occurring for much of this century. The real wage rate in agriculture increased an average of 1.45% a year between 1920 and 1979 (Hansen and Radwan, 1992; and Brunton, 1983). Furthermore, Aly and Grabowski found that the MPL in agriculture increased in the 1953-1972 period and actually became positive in the last two years. As a result of this transformation coupled with the migration caused by the oil boom, the surplus labor view of agriculture is much less relevant now than it was several decades ago.

Disguised unemployment in agriculture may have been replaced in importance by disguised unemployment in the public sector. Public sector disguised unemployment is suspected to exist in both the government bureaucracy and in publicly owned firms. Egypt is thought to have one of the largest public sectors of all the Third World economies, most of the country's industrial base being owned by the government. Government over regulation and an ancient bureaucracy has been thought to plague the public sector, to have held back technical modernization and to have reduced overall economic efficiency (CIA, 1995).

A suspected source of disguised unemployment in Egypt's public sector is the socialist policies initiated during Nasser's era (1952-1970) and the old governmental guarantee to secure a job for every college graduate. Recent reforms have sought to liberalize the economy and encourage foreign investment by reducing government bureaucracy, gradually abandoning job guarantees and privatizing the economy. However, the fear that disguised unemployment will be transferred into open unemployment, with all the political and social ramifications that might follow, has slowed the privatization efforts in some industries and stopped it in others. The textile industry is an example of the latter group of industries where the Egyptian government is only giving lip service to privatization.

Empirical Model, Data and Empirical Results

In discussing disguised unemployment in publicly owned Egyptian firms, first note that if the marginal product of labor exceeds subsistence, surplus labor does not exist. In our discussion, we will assume this to be the case. Disguised unemployment can still exist, although it will differ from disguised unemployment in the traditional sector in that there would be no link between the wage rate and the average product of labor. Public firms may be subsidized and, consequently, we could have labor's marginal product exceeding its average product. Hence, disguised unemployment implies that w > MPL but not that w = APL. Furthermore, it is possible that government owned firms over pay workers so that the wage rate is above the market wage rate. Hence, we can reject disguised unemployment as existing in publicly owned firms if their marginal product of labor exceeds the wage rate. If, however, w > MPL, we can only say that disguised unemployment might exist. With this cautionary note in mind, we will turn to an empirical examination of disguised unemployment.

In examining disguised unemployment in the public sector in Egypt, we will look at the textile industry. Besides the government's position mentioned earlier, with regard to the privatization of this industry, there are four reasons for choosing the textile industry. First, unlike government ministries, publicly owned firms have measurable outputs for estimating marginal products. Second, this industry has diversity in publicly owned firms in terms of size. Hence, we can see whether large publicly owned firms use more labor than small privately owned firms. Third, the textile industry uses relatively unskilled labor. Dualistic models of disguised unemployment assume unskilled labor. Fourth, this industry has a rich history to draw on since its inception in the first half of the nineteenth century by Muhammad Ali Basha (Wilson, 1995). We will examine the extent to which disguised unemployment does in fact exist in the textile industry.

The production function for textiles will be assumed to be Ray Homothetic, RH. An RH production function is an adaptable functional form, much more adaptable than the simple Cobb-Douglas, CES, Leontief or translog production functions. The RH function allows returns to scale to vary with output and factor intensity. To define the RH production function let x represent an input vector and

( x) = F{ ^{H(x/ x} G(x)}, (1)

where x represents the norm of x, > 0, and F is a monotonically increasing transformation of ^{H(x/ x} G(x) and G(x) has the property

G( x) = ^{H(x/ x )}G(x) (2)

with H(x/ x ) > 0. The function given by Equation (2) is a ray homogeneous production function. Thus, the ray homothetic function is a monotonic transformation of a ray homogeneous function. If the function H(x/ x ) is a positive constant for all values of x, it can be seen that equation (2) becomes a homogeneous production function and equation (1) a homothetic production function. Thus, the homothetic, homogeneous and ray homogeneous functions are special cases of equation (1).

The form of the RH production function to be estimated is

lnY = ln ₀ + ₁(L/A)lnL + ₂(K/A)lnK + ₃(M/A)lnM+ , (3)

where Y = real output, L = labor, K = capital and M = materials; A = L + K + M; ln _{0, 1, 2} and ₃ are the estimated coefficients; and is the stochastic error term. The marginal product of labor for this RH production function is given by

Y/ L = (1/A)Y{ ₁[1 + lnL - (L/A)lnL] - ₂(K/A)lnK - ₃(M/A)lnM}. (4)

The other marginal products can be similarly found.

The data used in the study were collected by the Ministry of Industry in Egypt and were derived from the annual general budgets and current accounts of all public textile firms (thirty in total) for fiscal year 1990. Real output, Y, is the total value of textiles produced and sold by each firm during 1990. L is the number of workers hired by each firm. K is the real value of total assets owned by each firm. Finally, M is the real value of all intermediate goods and services produced by each firm. M and K are expressed in terms of the labor costs of materials and capital.

The estimated equation is shown in Table I. Note that all the variables are significant and have the correct sign. Furthermore, R-square is a respectable 0.93. For comparison, the estimates for a Cobb-Douglas type model are also shown in Table I. Note that the coefficient for capital is the wrong sign, negative, in the Cobb-Douglas type model. Hence the RH production model gives theoretically acceptable estimates while the log-linear model (Cobb-Douglas) does not.

Given these results and the level of inputs for each firm, we can calculate the MPL for each firm. The MPL for each firm is shown in Table II. Note that none of these marginal products are below the subsistence wage rate of 1,204 Egyptian pounds. Hence, surplus labor, as defined above, does not exist in the Egyptian textile industry. The absence of surplus labor is not surprising given our previous discussion of the transformation of the Egyptian labor market.

There is evidence that disguised unemployment exists in the textile industry. Twenty-one of the 31 firms seem of have disguised unemployment (MPL < w), while only 10 firms use too little labor (w < MPL). Consequently, there is the possibility that chronic, disguised unemployment does exist in the Egyptian textile industry. Privatization of the textile industry could then result in turning this disguised unemployment into open unemployment.

The extent of this open unemployment would depend upon which firms are hiring too much labor. We might expect that the large firms hire too much labor because the hiring of large firms is more visible to policy makers. In addition, large firms could be closer to the center of power where open unemployment could be more of a concern to policy makers. To see whether disguised unemployment exists in the Egyptian textile industry as a whole, we need to look at the size of firms. There does not appear to be any particular pattern as to which firms hire too much labor. We can see in comparing Table II with Table III, that 75% of the 12 largest firms have disguised unemployment, while only 63% of the smaller firms have disguised unemployment. However, the two largest firms do not hire too much labor. The ten firms hiring too few workers, MPL > w, hire 100,714 workers. The 21 firms with too much labor, MPL < w, hire 133,401 workers. Consequently, the elimination of this implied misallocation of resources could, in the short run, result in a conversion of disguised unemployment into open unemployment.

Conclusion and Policy Implications

In the introduction it was pointed out that advocates of privatization argue that privatization would eliminate the inefficiency in the textile industry. We found evidence that this inefficiency does exist in that too much labor is employed, at least in the textile industry. If privatization would result in the competitive solution, where w = MPL, privatization would have three impacts on employment. First, the 10 firms where w < MPL would hire more workers. Second, the 21 firms where w > MPL would hire fewer workers. The net impact would be a minor, short-term decline in employment in the textile industry. Third, in the long-run, firms suffering losses might go out of business while firms making losses could expand. Again, the net impact, at least in the short run, would be some job loss.

Opponents of privatization, in Egypt and elsewhere, worry about the possibility of turning disguised unemployment into open unemployment. The results show that job loss could occur for the three reasons listed above. In the long run, new employment could be found elsewhere in that there is little evidence of surplus labor in the Egyptian economy as a whole. Furthermore, this study would seem to indicate that the disguised unemployment, resulting from w > MPL, could be minor.

The major concern with unemployment resulting from privatization is not that disguised unemployment exists but that excess capacity in the textile industry might exist. Of the twelve largest firms, only two make profits. The other eight profitable firms are among the twenty small firms. Some large firms would undoubtedly fail with privatization. Their management has an obvious reason to resist privatization and the management and workers at large firm often have political clout.

In discussing the unemployment of workers from privatization, it is important to correctly assess the possible source of this unemployment. If it is because disguised unemployment is converted into open unemployment, the subsidy of public sector firms might be viewed as a sort of unemployment policy. However, if the problem is simply inefficient firms in an industry where firms operate far below capacity, privatization will allow workers and capital to flow to sectors that do not have excess capacity. .

Adams, Richard "Worker's Remittances and Inequality in Rural Egypt," Economic Development and Cultural Change 35 (October, 1989):45-71.

Aly, Hassan Y. Belbase, Krishna and Kraft, Stephen, "The Technical Efficiency of Illinois Grain Farms: An Application of a Ray Homoghetic Production Function," Southern Journal of Agricultural Economics, 19 (July, 1987):69-78.

and Grabowski, Richard, "Technological Change and Surplus Labour in Egyptian Agriculture, 1952-1972," The Journal of Agricultural Economics 35 (January, 1984):109-15.

Bruton, Henry J. "Egypt's Development in the Seventies," Economic Development and Cultural Change, 31 (July, 1983):679-704.

Central Intelligence Agency, CIA World Facts Book: 1995. On line documentation, URL Addtess: http://www.odci.gov/cia/publications/95fact/pubinfo.html

Dar Akbar El Yom, Akhbar El Yom (a weekly news paper), issue No. 2655 (September 23, 1995):9. (In Arabic)

Färe, Rolf, Grosskopf, Shawna and Lovell, C. A. Knox, Measurement of Efficiency of Production (Boston: Kluwer Nijhoff, 1985).

Jansson, Leif and Lovell, C. A. Knox, "Modeling Scale Economies with Ray Homothetic Production Functions," Review of Economics and Statistics, 47 (November 1985):624-29.

Grabowski, Richard and Shields, Michael P. Development Economics (Cambridge, Mass: Blackwell, 1996).

Hansen, Bent and Radwan, Samir "Employment Planning in Egypt: An Insurance Policy for the Future," International Labor Review, 121 (September-october, 1982):535-51.

Ibrahim, S The New Arab Social Order, (Boulder: Westview Press, 1982).

Khalifah, F., "Economic and Social Impact of Egyptian Migrants on the Egyptian Economy: 1970-1983," Journal of the Social Sciences, 14 (1986) (In Arabic)

Koraym, K., "Economic Impact of Labor Migration on Rurual Egypt," Egypt Contemporary, 77 (1986) (In Arabic)

Lewis, W. Arthur "Economic Development with Unlimited Supplies of Labour," The Manchester School (May 1954), pp. 139-191.

Richards, Alan and Waterbury, J. A Political Economy of the Middle East (Boulder: Westview Press, 1990).

Sen, Amartya K. "Peasants and Dualism With or Without Surplus Labor," Journal of Political Economy, 74(1966):425-50.

. Employment, Technology and Development (Oxford: Clarendon Press, 1975).

Wilson, R. Economic Development in the Middle East, (London: Routledge, 1995).