Publications  Income Distribution


Empirical Evidence

from Poland and Czechoslovakia



It would be beyond the scope of this paper to attempt an exhaustive study of all aspects of income distribution. In any case, such a task could not be accomplished because the original source data on personal income in Poland and Czechoslovakia are not available to Western scholars. The only source of information available to us are data already processed by simple statistical routines, i.e. averages, frequency tables or cross-tabulations. For these reasons we shall concentrate our attention only on a few selected aspects of economic inequality, while other aspects may be mentioned only casually, if at all.


Although we have almost no data on the distribution of personal wealth in Poland and Czechoslovakia, we may quite safely conclude that this source of income inequality has been almost totally eliminated. This does not mean that wealth is distributed equally - we know that considerable inequalities in this area still persist - it simply means that most of the privately owned property is 'unproductive’ and as such cannot generate income. There are a few exceptions. such as private farming in Poland for instance, or interest paid on personal savings, but private farmers are not very rich, and the interest rate on personal savings is only nominal.

Poland and Czechoslovakia also seem to have very few problems with income inequality based on race and ethnicity, and this is of course mostly due to the fact that both countries are in this respect relatively homogeneous. There are some indications that Czechoslovakia's policy of assimilating the gypsy population has not been very effective, but unfortunately no data on the current situation of the gypsies are available.



Somewhat more serious is the discrepancy which still exists between incomes of Czechs and Slovaks. This is an outcome of the historically lower level of economic development in Slovakia, rather then of a deliberate discrimination of Slovaks. Actually the preferential treatment of Slovakia, especially in the allocation of investment, has led to an apparent closing of the gap between the two nationalities. For example, a comparison of average monthly wages in the two Republics11 shows that in 1955 Czech wages were 6.1 % higher than Slovak wages, in 1960 -3.4%, in 1970 - 1.9 %, and by 1974 only 1.2% higher. However, the comparison of average monthly wages may lead to an underestimation of income inequality between Czechs and Slovaks. The micro-census data (family budgets) indicate, that the per capita income in Slovakia is still almost 20 per cent lower than in Bohemia and Moravia.

Table 1  Distribution of Czechoslovak Households 
According to Net Money Income per Capita (%)

Click to display the table  ==>

in this window in a new window


There have been two remarkable tendencies in the intersectoral income distribution in postwar Czechoslovakia:

1. Income differentials among broad sectors of the economy have been diminished and the ranking of sectors has changed considerably 

Table 2 Sectoral Monthly Average Wages 
in Czechoslovakia

Click to display the table  ==>

in this window in a new window


2. Income differentials among branches of industry has not diminished and their ranking remained almost unchanged 

Table 3 Average Wages of Blue Collar Workers
 in Selected Branches of Czechoslovak Industry

Click to display the table  ==> in this window in a new window

Source: Statistical Yearbooks of CSSR, 1966, p.210. and 1972, p.250.


3. Another interesting feature of income distribution in Czechoslovakia is the drastic change which has occurred among the relative incomes of three basic categories of workers in industry 

Table 4 Average Wages of Main Categories 
of Workers in Czechoslovak Industry

  1948  1960 1970
(1) Blue collar workers 734  1406  1902
(2) Engineering-technical personnel  1214  1868 2569
(3) Administrative staff 914 1225 1626
Ratio 2:1 in % 165 133  135
Ratio 3:l in %  125 87 86

Source J. Adam op.cit, p.83.


There is very little direct statistical information about the relation between the level of education and incomes in Czechoslovakia. Data in Tables 2 and 4 seem to indicate that income differentials between more and less educated people have diminished, and in some cases the relation has been reversed. For example the average monthly earnings in Czechoslovakia in 1965 of a skilled coal-miner was 3521 Kcs, that of a lathe operator 2422, a doctor 2243, a locksmith 2010, a grammar school teacher 1907, a bricklayer 1865, and a hospital nurse 117812. Similarly, the cumulative income at age 60 in Czechoslovakia in the early 1960's (in thousands Kcs) was 3125 for an assembler, 989 for an engineer, 949 for a farmer, 900 for a technician, 888 for a lawyer, 771 for an economist, and 692 for an unskilled worker13


Probably the most striking feature of income distribution in Czechoslovakia is the persistent discrepancy between the wages and salaries of men and women. Notwithstanding the facts that (1) Marxist ideology clearly condemns discrimination of women, (2) Czechoslovak law gives them the legal right to the same wages as men, (3) the excess demand for labor and governmental policies of assistance to working women (e.g. pregnancy leaves, subsidized day care centers etc.) has led to a very high level of women's participation in the labor force (47 % in 1974), and (4) equal educational opportunities are available to men and women, women nevertheless receive on average only two-thirds of the average income of men.


Table 5 Average Monthly income of 
Men and Women in Czechoslovakia

May of the Respective Year in Kcs socialist sector industry
1959 1968 1970 1959 1968
1) men 1596  2106 2338 1689 2140
2) women 1046  1140 1565  1054 1350
 Ratio 2.1 in % 65.5 66.5 66.9 62.4 63.3

 Source: J. Adam, op.cit., p. 87.


In the breakdown of wages by industries  we see that the overall discrepancy between the incomes of men and women can be explained partly by the fact that the women's participation ratio is higher in sectors with lower wages, and partly by the fact that in each sector the wages of women are lower than the wages of men.

Table 6 Monthly Average Wages 
in Sectors of Czechoslovak Economy

Click to display the table  => in this window in a new window


Several facts are noticeable: First, there is a clear negative correlation between the women's participation ratio and the level of wages among sectors. Second. the sectoral differences in women's wages are much smaller than the differences in men's wages. The standard deviation of the women's wage is only 63.8 Kcs while the standard deviation of men's wage is 166.5 Kcs. Third, the sectors with the highest women's participation level (health. education and trade) are those which have lost most in terms of relative wages (see Table 2), whereas the sectors with the lowest women's participation are those which have gained. Fourth, the above also holds for differentials among branches of industry. It can be shown that the branches with the highest women's participation level are those where the wages are lowest (e.g. textile, food), while the women's participation level in the branches with highest wages (e.g. metallurgy and coal mining) is very low. These facts imply that in spite of a considerable increase in women's participation in the labor force, which was also accompanied by an increased level of education and training, the gap between wages and salaries of males and females has not diminished. Can this be explained rationally from the principles of socialist income distribution? Most likely not.


It may be argued. that the increased school enrollments and the increased employment of women are relatively recent phenomena. so that their average level of education and work experience is still lower than the levels of education and experience of men. This may be true, but are the differences in education and experience large enough to explain the entire gap between incomes of men and women, so that pure sex discrimination can be ruled out? To answer this questions we need data which would allow us to estimate simultaneously the role of education, experience and sex in the determination of personal incomes. Unfortunately, this kind of data cannot be obtained for Czechoslovakia. However, Polish statistics14 contain the following two types of data, which can be used for our purpose:

  1. The number of fully employed persons in the socialist sector of Polish economy in a 4-way breakdown according to 22 administrative regions (wojewodztw), 13 economic sectors, 5 levels of education, and sex.
    Four categories of wage funds and the total number of employees in a 2-way breakdown according to 22 administrative regions (wojewodztw) and 13 economic sectors


Both sets of data are available for the years 1970 and 1971. The second set of data allows us to calculate average wages for each sector of each region for both years, by simply dividing the wage fund by the number of employees in the respective cells of the cross-tabulation. Two types of average wages will be used in the following regressions as left-hand variables.

  • W1 - average wages of type I include only payments from the so called 'Personal wage fund' -

  • W2 - average wages of type 2 include (in addition to W1) all other payments to individuals from the so-called ‘nonpersonal wage funds', and other funds of enterprises. This would include bonuses, honoraria, per diem reimbursement of travel cost, etc.


The first data set was used to calculate the shares of individual education levels, as well as the women's participation ratio in each sector of each region for both years.

The following variables were thus obtained

UN ... share of employees with university education

SP ... share of employees with secondary professional education

SG ... share of employees with secondary general education

EP ... share of employees with elementary education and professional (vocational) training

WOM.. women's participation ratio


Our main task now will be to check whether, and how much of the sectoral and regional variation of average wages can be explained by the above defined five explanatory variables. We shall therefore attempt to estimate the following regression equation

W = b + b1UN + b2SP + b3SG + b4EP + b5W0M + e

In the actual run several alternatively defined dummy variables and a special corrective variable which was designed to capture the impact of differences in the definition of total employment in the two sets of data were also included. The coefficients of this corrective variable as well as those of the dummy variables and the constants were in almost all cases significant, but they are not very interesting and therefore will not be reported.


Table 7

Click to display the table  ==> in this window in a new window


The first regressions were run for each region separately across sectors. To increase the number of observations, data for the two years were pooled and a dummy variable for 1971 was included in the regressions. The total number of observations was 22 for each of the five cities, and 26 for each of the remaining regions so that the residual number of degrees of freedom is relatively small: 35 for cities and 19 for the other regions. Although the regressions were run for both types of wage variables, only those for W1 are reported, because W2 gave the same kind of results.


The results of Table 7 show:

  1. A relatively high share (.7 - .9) of sectoral differences of average wages in each region are explained by the explanatory variables of the model.
  2. Consistently, the most significant explanatory variable is the women's participation ratio. The negative coefficients of WOM indicate that the average wage is 100 - 200 zlotys lower for each percent of women employed in the sector. From this type of regression however, we cannot conclude whether the lower average wage is due to the fact that women have lower wages than men, or to the fact that wages, including the wages of men, are generally lower in sectors with higher women participation. Probably both are true.

  3. The majority of coefficients of educational variables are positive, but they vary quite a lot and are frequently insignificant. The only safe conclusion from these regressions seems to be that the average wages in sectors are positively related to the percentage of employees with a higher than elementary education. It cannot be conclusively established, however, that a secondary education will yield higher wages than an elementary education with vocational training, or that a university education always results in a higher income than a secondary education.


Table 8

Click to display the table  ==> in this window in a new window


The second set of regressions (see Table 8) was run for each sector separately and across regions. Again, data for both years were pooled, and a dummy variable for 1971 was included in the equation. In each of the 33 sectoral regressions there were 44 observations so that the residual number of degrees of freedom was 36. This time, results for the left-hand variable W2 are reported; the regressions with W1 on the left-hand side gave very similar results. Although the regression equation in this case contains the same types of variables as the first case, the meaning of the regression coefficients is not identical. For example the coefficient of WOM will no longer contain the effect of the fact that in sectors with higher women's participation, wages of men may be lower. It may seem at first glance, that the regressions should give better results because they include more homogenous units. This however, is not necessarily true, because the results may be biased by the fact that in regions with a higher general level of wages (i.e. in cities) is concentrated a higher proportion of managerial, clerical and research staff.

The results in Table 8 do actually present a somewhat different picture than the results in Table 7. With the sole exception of industry, the coefficients of WOM are small, insignificant and sometimes have positive signs. The estimated coefficients of the educational variables also seem to be unreliable - they frequently take on strange values and signs and are very rarely significant. Only UN (the university education variable) has consistently positive, very high and rather significant coefficients.


Table 9 reports results of regressions which were run on pooled regional an sectoral data. Because pooling guaranteed a sufficient number of observations (276), it was possible to run the regression for each year separately. Two sets of dummy variables were introduced to control for the effects of regional an sectoral differences in the wage levels, therefore the coefficients reported in the table should in the clearest possible way represent the contributions of different levels of education, and pure sexual discrimination to the wage level. It is interesting to see, that university education seems to have quite a sizable and clearly significant positive impact on wages. Secondary professional education also has positive, but significantly lower effect than university education. Elementary professional education apparently contributes not much less than secondary professional education to wages, while secondary general education has no recognizable effect on the level of wages. Finally, the large negative and significant coefficients of WOM indicate that women who have the same level of education as men are paid less.


Table 9



































































A summary review of the contribution of individual factors to the explained variation in average wages is given in analysis of covariance tables (Tables 10  and Table 11) which were constructed from the regression Table 9. (year 1970 only). The sets of variables were added successively in hierarchical manner, which may overstate the explanatory power of those which entered earlier and understate the explanatory power of those which enter later. Nevertheless even the last entering factor (regions) appears to be highly significant.

It is interesting that education (when entered first) explains 40 - 45 percent variation in both W1 and W2 and that sex alone (after controlling for five educational categories) explains an additional 16 percent of variation in W1, but only 7 percent of variation in W2. After controlling for education and sex, and correcting for the discrepancy in data, sectoral differences in average wage account for about 15 percent of variation in W1 and 25 percent of variation in W2. Finally 7 to 8 percent of variation in both W1 and W2 is attributable to remaining regional differences in average wages.


Table 10 Analysis of Covariance Table for W1 and 1970

Source of variation


Sum of Squares

Mean Square

F statistic

Education 4

2.2030 E9

550.75 E6

Sex 1 .8591 E9 859.10 E6 363.35
Education and sex


3.0621 E9

612.42 E6 249.02
Corrective variable 1 .6609 E9 660.09 E6 279.52


81098 E9





.373068 E9


Sub total 39 4.90705 E9 125.82 E6 53.22


.55799 E9

2.36 E6



5.465 E9

19.87 E6



Table 11 Analysis of Covariance Table for W2 and 1970
Source of variation


Sum of Squares

Mean Square

F statistic

Education 4

4.6789 E9

1169.6 E6

Sex 1 .7142 E9 714.2 E6 206.24
Education and sex


5.3926 E9

1078.5 E6 311.43
Corrective variable 1 .6551 E9 655.1 E6 189.17


2.4619 E9

205.2 E6




.6152 E9

29.3 E6

Sub total 39  9.1248 E9 233.9 E6 67.562


.81737 E9

3.463 E6



9.942 E9

36.15 E6




This study attempted to identify some of the primary factors which determine personal income distribution under Soviet-type socialism, and to compare the reality of the distribution with the normative statements of Marxian economic theory. The empirical evidence was based on scattered (and not always consistent) data for Czechoslovakia and Poland.

In conformity with Marxian theory, we find that income inequality has diminished and that wealth has ceased to be an important source of income differentials (this conclusion is based on evidence presented elsewhere). We also find that income inequality based on ethnic and regional differences has been diminishing since World War II, although some differences in personal incomes among Czechs and Slovaks, and among regions (wojewodztwa) in Poland, still persist.


It was shown in the first part of this paper that income differentials based on education (human capital) are considered by Marxists to be healthy and necessary for socialism. The evidence from Czechoslovakia seems to indicate that the role of education as a source of income differentials has diminished, and in some cases was reversed. The evidence from Poland, however, shows that education - primarily university education - is an important source of income differentials.

The empirical data demonstrate that considerable sectoral differences exist both in Czechoslovakia and Poland. This can hardly be justified in light of the Marxian normative theory of income distribution. However, the most striking conclusion is the fact that pure sex discrimination still remains as a major source of income inequality under Soviet-type socialism. This phenomenon is in clear contradiction with the normative Marxian view on income distribution.




OK Economics was designed and it is maintained by Oldrich Kyn.
To send me a message, please use one of the following addresses: ---

This website contains the following sections:

General  Economics:

Economic Systems:

Money and Banking:

Past students:

Czech Republic

Kyn’s Publications

 American education

free hit counters
Nutrisystem Diet Coupons