1. The socialized sector

We assume that labor is the only factor of production in the SS enterprise: ^1/ one unit of labor is required to produce one unit of the final product (an assumption to be relaxed later). We will suppose that the socialized enterprise is labor managed (or at least labor dominated) ^2/ and thus that its objectives can be represented by a utility function that depends positively on the employment and the real wage. ^3/

For simplicity we adopt a modified Stone-Geary form for this utility function:

where $\overline{w}$ is the private sector wage, w is the wage in the socialized sector and x is output (equal to employment) in the socialized sector. Expression (1) implies that the firm is interested in the after tax excess wage and employment. ^4/ The weight on employment in the firm’s objective function is intended to capture the tendency for “labor hoarding” mentioned in the introduction. Parameters θ ≥ 0 and γ ≥ 0 correspond to the excess wage and employment elasticities of U(.); we define the firm as “wage (employment) oriented” if θ>γ (θ<γ). We will assume that the SS enterprise is employment oriented for the rest of the analysis.

The utility function of the representative consumer, which is also the social welfare function, is specified as:

where S′(.) > 0, S″(.) < 0, x Is the production of the state enterprise, y is the amount of imports, and z is consumption of the private sector good. That is, we assume without loss of generality that V(.) is linear in z, and also assume that y and x are perfect substitutes.

Solving the consumer maximization problem, and denoting by P the relative price of good x in terms of good z, we express the inverse demand function of the socialized sector as:

The absence of an effective personal income tax or value added tax system, as discussed in the introduction, implies that the majority of government revenues are obtained through taxes on enterprises. A prevalent form of such taxes in MPEs is the turnover tax; in practice, the calculation of the turnover tax is quite complicated, but here we shall represent it as equivalent to unit sales tax. ^1/ The government imposes a unit tax equal to r = P(x+y) - w, where w is the wage negotiated between the firm and the government. We will assume that the government is interested in maximizing total tax revenues collected from the labor managed firm. These revenues expressed in units of good z are:

The labor managed firm and the government bargain over the wage w and employment x. This encapsulates the array of bargaining discussed in the introduction; since the government can influence output via its control of raw materials, credit and foreign exchange, and can influence wages (including bonuses) via the tax system, income policies, real interest rates and other instruments, both wages and output/employment can be regarded as determined by bargaining between the enterprise and government. The bargaining process assumed is a generalized Nash bargaining game (Nash, 1950). We assume that bargaining is conditioned on the wage in the NSS, $\overline{w}$, and on the level of imports y competing with the good produced in the SS. The level of imports is given by commercial policy, reflecting the prevalence of nontariff barriers in Eastern European countries. We shall assume for simplicity that commercial policy is not subject to bargaining between the government and the socialized enterprise. This framework is equivalent to assuming a government consisting of two independent agencies, one collecting revenues from state enterprises and one controlling imports. The Ministry of Trade imposes a binding import quota y, buys imports at international price and sells the quantity imported at the domestic price (P(x+y)). This approach permits an examination of the implications of alternative degrees of liberalization of the trade regime. ^1/

Based on these assumptions, the Nash solution maximizes the generalized Nash product defined as:

where 0 ≤ α ≤ 1 is a parameter capturing the relative bargaining power of the enterprise vis-à-vis the government.

The bargaining framework assumes that either the enterprise or government could shut down production in the SS if an agreement on wages and output were not reached; in this case output, the utility of the enterprise, and tax revenue would all be zero. Such a shutdown could be visualized as a strike by workers in the socialized enterprise, as strikes are infrequent, but certainly not nonexistent in socialist economies, but this interpretation would be too narrow: even in the absence of the strike instrument, the cooperation of the management and workers in the SS is required to generate any output, as output could be reduced to very low levels through shirking, sloppy work, or sabotage. In a labor-dominated firm, there are relatively few constraints on such actions. The government, for its part, could also bring output to zero, through its control of credit, foreign exchange and vital raw materials. However, the Nash bargaining solution implies that shutdowns do not occur; rather, agreement is reached at a wage such that the product (5) is maximized.

Maximizing the generalized Nash product G(.) with respect to the negotiated wage w and output x yields the following first-order conditions:

where subscripts and primes denote partial derivatives. Equations (5) and (6) determine the negotiated wage and output target in the socialized sector for a given level of imports y and for a given NSS wage $\overline{w}$.

2. The nonsocialized sector

The nonsocialized sector is characterized by perfect competition and constant return to scale. In addition to labor, there is a sector-specific factor T (say land); the sector’s technology is summarized by the following production function:

where Z(.) is a monotonically increasing quasi-concave function. Here z is consumption and z^* exports of the good produced, l the amount of labor employed in the NSS and φ is a productivity parameter. The wage in the NSS, $\overline{w}$, is determined competitively, and thus equals the marginal product of labor in that

where $\overline{w}$ is expressed in units of the good produced in the NSS, z, and $Z_l=\frac{\partial Z\left(\mathrm{.}\right)}{\partial l}\mathrm{.}$

Assuming that there is no unemployment, ^1/ the full employment of labor condition is expressed as:

where x is the number of workers employed in the SS (equal to that sector’s output), l is the number of workers employed by the private sector and L is the labor endowment.

3. International trade

Trade is modeled by assuming that the government collects the quota revenues, which are defined as:

where P^* is the international price of product x and y is the import quota. We may visualize the difference between domestic and world prices as similar to a specific tariff.

The model closes with the balance of payments condition:

which requires that the value of exports z is equal to the value of imports yP* plus external debt service E. ^2/

1. Trade liberalization

Trade liberalization is one of the major areas of market-oriented reform being carried out in socialist economies (see e.g., Wolf, 1989; and Plowiec, 1988). This section explores how protection from import competition affects the ability of the government to collect tax revenues which in turn can be used to pay external debt services. Appendix I contains the necessary algebra.

The government of the modified planned economy in this model has two sources of revenue, revenues from trade restrictions and revenues from state enterprises. ^1/ Both components of revenue are affected by the level of protection of the SS (i.e., the differential between domestic and world prices). Total revenues are used to finance a given level of domestic spending plus debt servicing E.

The revenue from trade restrictions (expressed in units of the NSS good z) is:

The change in R^y with respect to imports y is $\mathrm{d}{\mathrm{R}}^{\mathrm{y}}\mathrm{/}\mathrm{d}\mathrm{y}\mathrm{=}\mathrm{(}\mathrm{p}\mathrm{-}{\mathrm{p}}^{\mathrm{*}}\mathrm{)}\mathrm{+}{\mathrm{P}}^{\mathrm{\prime }}\begin{array}{c}\mathrm{<}\\ \mathrm{>}\end{array}\mathrm{0.}$

Figure 2 illustrates the standard relationship between the revenue from trade restrictions and the quota. Revenue is maximized at y_R, whereas it is zero at free trade imports y_FT and where imports are completely excluded. For imports y ≥ y_F the quota revenue is zero.

TRADE LIBERALIZATION AND REVENUES

Citation: IMF Working Papers 1991, 075; 10.5089/9781451849639.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

TRADE LIBERALIZATION AND REVENUES

Citation: IMF Working Papers 1991, 075; 10.5089/9781451849639.001.A001

• Download Figure
• Download figure as PowerPoint slide

TRADE LIBERALIZATION AND REVENUES

Citation: IMF Working Papers 1991, 075; 10.5089/9781451849639.001.A001

• Download Figure
• Download figure as PowerPoint slide

The second source of revenue consists of enterprise taxation π:

Differentiating ∏ totally with respect to the quota y,

The first term in brackets shows the increase in tax revenues for a given output x, because a change in protection alters both the domestic price and the negotiated wage. The second term shows the effect of protection on tax revenues for a given price and wage, through the change in output. Notice that dx/dy < 0 because of (A2) and that dx/dy + 1 - [(1-b) P′+φZ_ll]/D > 0. The change in the negotiated wage due to protection can be-obtained by differentiating totally the equation of the Nash bargaining curve and substituting the appropriate expressions from Appendix I:

The negotiated wage decreases in the amount of imports because a higher level of imports reduces the demand for labor and output in the socialized sector. Substituting dx/dy + 1, dw/dy, and (20) into (18), we obtain

where D < 0 and dx/dy < 0. A sufficient condition for revenues π to increase in imports y that the term in brackets is positive. However,

where $d\overline{w}/dy=-\phi Z_{ll}\frac{dx}{dy}<0.$

In other words, trade liberalization decreases both the negotiated wage and the private sector wage. However, when trade restrictions have a smaller effect on the NSS wage than on the SS wage (i.e.,| dw/dy | ≥ | d$\overline{w}$/dy |), tax revenues collected from state enter-prises decrease with trade liberalization. This would also be true if the marginal product of labor in the NSS is constant (φZ_ll=0) or if demand in the socialized sector is linear (P″=0). In order to proceed with the analysis, we shall assume that tax revenues π decrease with trade liberalization.

Figure 2 illustrates the total revenue as well as the state enterprise revenue as a function of the price differential. Free trade results in revenue II_FT because the revenue from trade restrictions is zero. Curve BD shows the revenue collected from state enterprises, which increases monotonically until it reaches the level 0D where there is no trade. Total revenue remains fixed for any amount of imports greater than y_FT. In Figure 2 a level of imports y_o maximizes the level of revenue available to the government; y_o is less than the maximum trade revenue quota y_R. Curve BCD shows the total revenue collected by the government, π+R, as a function of the level of permissible imports. This is the revenue available for debt service payments given the institutional features of a reformed socialist economy.

This analysis implies that in a socialist economy, trade liberalization reduces revenues in two ways: it reduces the trade revenue, and, by reducing the market share of the monopoly producer, it reduces the revenues that can be collected through taxation of the socialist enterprise. This implies that, given the existing fiscal structure, the government’s revenue requirements affect the extent to which trade liberalization can take place. For instance, in Figure 2, if the initial revenue requirement is R₁, then the level of imports cannot increase above 0y₁. If debt reduction takes place, reducing the government’s revenue requirement to R², further trade liberalization can take place, increasing the level of imports to 0y₂. This illustrates the way in which debt reduction can ease the fiscal constraint on trade liberalization.

2. Demonopolization

Another important set of reforms planned or begun for many socialist economies is demonopolization. The pervasiveness of monopolies in socialist economies is generally viewed as a source of inefficiency, and breaking up monopolies is accordingly regarded as an important goal of market-oriented reform.

Demonopolization can be analyzed in the model as follows. Suppose that the government breaks up the monopoly in the socialized sector into firms competing in the product market. We shall assume that these firms’ behavior is characterized by the Cournot model, such that each takes the others’ output as given when making its own output decision. We shall also assume that there is no overarching institutional structure, such as an industry-wide trade union, that enables the enterprises to set wage and employment for the industry as a whole; rather, we shall assume that each enterprise bargains with the government independently over the tax that each will pay (and therefore the residual available to the workers in wages and bonuses), given the outcome of bargaining between the government and the other enterprises. In considering the effects of competition policy, we shall also assume a given quota establishing the level of imports y; this is tantamount to assuming that the Ministry of Trade is another player in the Cournot game played by the enterprises. As in the previous analysis, it is assumed that trade protection contributes to the revenues collected by the government.

Given these assumptions, the objective of the i’th firm is

The revenue collected from each enterprise is then

Then, supposing that bargaining is carried out separately between the fiscal authorities and the management of each enterprise, the bargaining solution maximizes the generalized Nash product

Then assuming for simplicity that all firms are symmetric, the condition determining the firm’s employment and wage is

where condition (26) is therefore analogous to condition (15) for the monopolist.

By differentiating condition (26) totally, the effect of increased competition and imports on the output and employment of the SS, x, can be determined:

Condition (27) shows that, as one would expect, breaking up the monopolist in the SS into more pieces (increasing n) reduces the output of each resulting firm, but condition (28) shows that it results in an increase in the overall output of the SS. This result implies that an increase in competition reduces the price and increases consumption of the good produced in the SS; because it shifts labor into the SS from the NSS, it therefore increases the equilibrium wage in the NSS. The wage resulting from negotiation between the enterprise and the government is therefore also affected by competition:

Increased competition in the SS thus has two opposing effects: the direct effect of the competition is to reduce excess wages in the SS, but by shifting labor into the SS, it also puts upward pressure on the wage in the NSS. A sufficient condition for the direct effect to dominate, leading the sign of (29) to be negative, is that |φZ_ll| is not large; under this condition, increased competition in the SS reduces the wage that labor-managed firms there are able to obtain for their employees. As the number of firms becomes large, the wage gap w-$\overline{w}$ approaches zero.

Relaxing trade restrictions has a similar effect under Cournot oligopoly as it was shown in the previous section to have under monopoly: it leads to a reduction in the output of each enterprise in the SS, and thus to a reduction in the SS’s output:

Finally, the effects of demonopolization on the profits (net of bonuses to the workers) of the SS, which constitute a major source of revenues for the authorities, can be determined. Denoting these revenues by π = -(1-κ)x²nP′ (nx+y) and differentiating with respect to the number of firms n into which the monopoly is broken, we obtain

Thus, increased competition in the SS generally reduces the revenues obtained from that sector.

We can also derive the effects of demonopolization on the revenues obtained from trade restrictions. Since the revenues from a quota depend on the difference between the domestic and foreign price, these revenues R are reduced if domestic output is increased and the domestic price is reduced:

Thus, increased competition in the domestic market reduces the rents obtainable via trade restrictions.

Demonopolization of the SS therefore decreases government revenues for two reasons: the revenues obtained from the profits of the socialized enterprises decrease, and so do the revenues associated with trade restrictions. These revenues are lower, the greater the degree of competition created in the socialized industry, represented in this model by the number of pieces into which the socialized enterprise is split. This relationship is illustrated in Figure 3. If there is a minimum required level of revenues R₀, this implies that a maximum number of firms n₀ can be allowed to compete. Thus, to the extent that the government of a MPE is constrained to obtain a particular level of revenues, for instance for servicing its external debt, it will be constrained in its ability to demonopolize the socialized sector--perhaps creating profitable oligopolies rather than going all the way to establishing competition.

DE-MONOPOLIZATION AND TAX REVENUES

Citation: IMF Working Papers 1991, 075; 10.5089/9781451849639.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

DE-MONOPOLIZATION AND TAX REVENUES

Citation: IMF Working Papers 1991, 075; 10.5089/9781451849639.001.A001

• Download Figure
• Download figure as PowerPoint slide

DE-MONOPOLIZATION AND TAX REVENUES

Citation: IMF Working Papers 1991, 075; 10.5089/9781451849639.001.A001

• Download Figure
• Download figure as PowerPoint slide

3. Technical improvements

In establishing priorities for the transitional socialist economies of Eastern Europe, much attention has been given to measures intended to increase productivity. Some such measures are related to the restructuring of enterprises; it is held that many of the socialized enterprises suffer from poorly-designed internal structures, diseconomies of scale, inappropriate product mix, and antiquated management procedures. Some measures have more to do with the need for investment to replace obsolete equipment and provide better infrastructure. Others relate to the transfer of technology, both with regard to more efficient production techniques and to more effective management practices. Although the emphasis has been on increasing the efficiency of the socialized sector, either before or after privatization of socialized enterprises, there have also been some efforts directed toward the nonsocialized sector, as seen for instance in the Task Force on Polish Agriculture established in 1990, sponsored jointly by the World Bank, the European Community, and the Polish Government.

In this section, we shall consider the implications of improvements in productivity, in the context of our simple general equilibrium model, with regard to their implications for the distribution of output between the SS and NSS and for the revenue-raising capacity of the government under the assumption that there is only one firm in the SS. Here, we shall not be concerned with the origins of the productivity improvement, as our model does not distinguish between improvements associated with restructuring, investment, or technology transfer. We shall therefore consider the implications of technical improvements, showing how these implications differ depending on the sector in which they take place.

Assume that each worker hired by the state enterprise produces ψ units of output (instead on one) and labor is the only factor of production. Consequently, total output in the SS is ψx where x denotes the number of workers in the SS as before. The equilibrium condition of the model is given by:

where P(ψx+y) is the domestic price in the SS.

By totally differentiating (33) we can calculate the change in employment x due to changes in productivity in the SS and in the NSS. The impact of a technical improvement in the SS ψ is given by:

Consequently an increase in labor productivity in the SS decreases employment in the state enterprise and increases employment in the NSS. However, total output in the SS increases as a result of higher sectoral productivity.

Equation (35) implies that total consumption y+ψx in the SS increases and the domestic price declines. Moreover, because employment in the NSS increases, its output rises and the wage $\overline{w}$ is reduced when productivity ψ increases.

Higher productivity in the SS also raises profits in that sector; this follows from the fact that in the bargaining equilibrium marginal revenue exceeds the outside wage $\overline{w}$.

Now consider a technical improvement in the NSS:

An increase in productivity in the NSS decreases employment in the SS sector and reduces output in that sector; it raises the domestic price P and the NSS wage $\overline{w}$. Technical improvement in the NSS may therefore lower profits in the socialized enterprise.

4. Bargaining power

An important aspect of the model is the bargaining power of the labor-managed socialized enterprise, in extracting as large as possible a share of the monopoly rents, in terms of a combination of gains in wages and employment, in the socialized industry. One can consider institutional reforms that would influence the bargaining power of the enterprises vis-à-vis the government. Some examples of such reforms might include (a) the introduction of tough wage control laws limiting wage increases on a uniform basis for all enterprises, and thus reducing the enterprises’ ability to bargain for higher wages; (b) increasing the uniformity, and stiffening the enforcement, of tax laws; (c) restricting the activities of trade unions, which might otherwise provide an additional institutional channel for demanding higher wages; (d) setting tight budgetary limits on subsidization of state enterprises; (e) setting ceilings for the expansion of credit to the SS. Such policy changes might be viewed either as changes in the framework in which bargaining between enterprises and government takes place, or simply as moves in the bargaining game; they might actually change the parties’ bargaining power if the government made a commitment to change policies in these ways, particularly if this commitment were entrenched as part of an agreement, for instance as part of a program of structural reforms on which international lending was made conditional.

Changes in the bargaining power of the enterprise can readily be analyzed as a change in the parameter α in the model. For the simple model developed in Section 2, increasing the bargaining power of the enterprise γ increases employment in the SS. (See equation (A3) in Appendix I), which implies a lower P and a higher $\overline{w}$.

5. Privatization

The most visible and controversial aspect of reform planned for socialist economies is the privatization of state enterprises. It is also the most difficult to treat in a general equilibrium framework, for the following reason: it is not the titular ownership of the means of production itself, but the more specific allocation of property rights and the mechanisms whereby the new privatized enterprises are controlled, that is important (e.g., Frydman and Rapaczynski, 1990).

The implications of privatization may reflect a combination of several changes that have already been considered. Privatization may include an element of demonopolization, as has been the case, for example, with much of the piecemeal “spontaneous privatization” that has taken place in several Eastern European countries to date, as well as for instance with the privatization of retail outlets that took place in Poland in 1990. This is not a necessary concomitant of privatization, however: particularly if privatization takes place through sales, and if the government is short of revenue, there is a strong motive for the government to leave an enterprise’s monopoly position intact, in order to increase its market value. ^1/ To the extent that demonopolization is a result of privatization, it would have effects such as are analyzed in subsection 2.

A second possible implication of privatization is a restructuring of the enterprise, increasing productivity. This would occur to the extent that privatization results in the management’s being chosen and influenced by a group that has a stake in the enterprise’s long-term profitability; such a group would set up more effective management procedures to provide greater incentives for effort as well as better coordination of activities. To the extent that this results from privatization, it would be reflected in an increase in productivity such as is analyzed in subsection 3.

A third possible implication of privatization is a change in the government’s bargaining power vis-à-vis other claimants on the income streams generated by the enterprise. To the extent that this occurred, its effects would be as analyzed in subsection 4. This would occur, for example, to the extent that privatization entails taking property rights from the workers, as represented by the Workers’ Councils, and giving (or selling) them to the new owners. In this case, if there were competition among the set of potential owners, bringing the owners’ expected excess return to zero, the government’s bargaining power to obtain a larger share of revenues from the enterprises would be increased through privatization. On the other hand, a lack of competition among potential owners might give the owners bargaining power at the expense of both the workers and the government; in this case, the net effect on the government’s bargaining power might be either positive or negative. The workers’ ability to bargain over wages via the enterprise councils might also, to some extent, be replaced by the increasing activity of trade unions.

Privatization may also have direct implications for the government’s revenue constraints. First, privatization may have a direct effect on the time profile of revenues, to the extent that part or all of the shares in privatized companies are sold. If shares of the privatized enterprises were given away through a voucher system or some other method, rather than being sold (e.g., Sachs and Lipton, 1990; Frydman and Rapaczynski, 1990), this effect would of course be nullified. Secondly, to the extent that revenues would have to be extracted from privatized enterprises through a uniform, and potentially distortionary tax system rather than from each enterprise through case-by-case bargaining, this would reduce the revenues obtainable from the enterprises.

Therefore, privatization is likely to affect the revenues available to the government, increasing these revenues to the extent that it results in technical improvement or decreases the bargaining power of the workers, but decreasing it to the extent that it results in demonopolization or to the extent that it creates a group of owners with bargaining power at the expense of that of the government. These effects are in addition to the more straightforward revenue consequences of privatization, which depend on the specifics of the arrangements whereby privatization takes place. Therefore, the government’s revenue constraint may dictate either a faster or slower speed of privatization, and may also suggest arrangements that would be more likely to enhance the revenue obtained by the government.