Property confiscation and the intergenerational transmission of education in post-1948 Eastern Europe

Abstract

Using regression methods and propensity score matching applied to two different retrospective samples, this study finds evidence of a positive “property confiscation” effect on educational attainment. We use a 1993 survey of adults (aged 20–69) in the post-transition Eastern European countries of Bulgaria, the Czech Republic, Hungary, Poland and Slovakia. In countries experiencing the most private property losses, regression results indicate that years of schooling increase by about 0.19 for each member of an affected extended family (parents, maternal grandfathers, or paternal grandfathers). When all three sets of family members lost property, we find an increase in years of educational attainment of about 0.6. We also find an increase in the probability of post-high school education of about 0.02 for each extended family member whose property was confiscated. Those findings are confirmed using propensity score matching, which provides a larger and more pervasive positive confiscation effect. We also test our hypothesis using current and retrospective microeconomic panel data from Europe’s Survey of Health, Aging and Retirement (SHARE), a dataset that covers countries in Eastern and Western Europe. We again find that property confiscation leads to greater educational attainment in the children of the affected households. We apply propensity score matching to the data and find, again, positive and statistically significant evidence of a confiscation effect on years of educational attainment. Auxiliary work indicates a separate channel for property confiscation’s effects. Our explanation for the empirical results reported herein can be found in families’ ability to pay bribes to advance their children’s education.

Appendices

Appendix 1: Two-period model of the impact of property confiscation

Below we develop and solve the problem of a utility-maximizing representative household willing to provide education to their offspring in a socialist country. Parents earn income (from unskilled and/or skilled labor), pay taxes, spend on consumption, and may have their property confiscated. Parents also save and bequest, and both are subject to seizure and taxation.

The model has a similar structure as the bequest models of Araújo and Martins (1999), Faria and Wu (2012) and the intergenerational model of Gandelman (2016). Gandelman (2016) analyzes the case in which, with uncertain house property rights, parents may invest in children's education or in their own homes. He derives and analyzes income and substitution effects. We also address similar kinds of choices in our model, however the trade-off involves education of present versus future generations, under the ever-present threat of property confiscation.

In line with the literature (e.g., Checchi 2006), the representative household maximizes a two-period utility function that takes into account parents’ and children's education and reflects bequest motives under the risk of property confiscation. The model yields a number of important results. First, it is consistent with two stylized facts of a socialist economy: the gradual erosion of the standard of living and the ever-present risk of famine. In addition, it shows that increases in bequest confiscation increase the education of parents and reduce the education of children. Finally, it shows that education of parents and children are complements, and that confiscation of capital amplifies this effect.

The model notation is as follows: C = consumption, K = capital, H = education, L = labor, σ = tax on bequests, r = interest rate, δ = discount factor, η = tax on capital, G = government transfers, and p = relative price of education.

In our construct, education is a good that increases utility

$$Max U\left( {C_{t} ,C_{t + 1} ,L_{t} ,H_{t} ,H_{t + 1} } \right) = \ln C_{t} + \delta \ln C_{t + 1} - \lambda \left( {L_{t} } \right) + v\left( {H_{t} } \right) + V\left( {H_{t + 1} } \right)$$

(3)

with $\nu ', V', \lambda ' > 0; \nu '',\, V''\left\langle {0,\, \lambda ''} \right\rangle 0$.

The utility maximization is subject to the budget constraints for periods t and t + 1. In period t, parents consume, $[C_{t} ],$ and invest in their own education, $[p_{t} H_{t} ]$. Their sources of income are unskilled labor earnings, [$w_{t} L_{t} ]$, and the return on capital inherited from a previous generation that was not confiscated, but was heavily taxed, [$r_{t} \left( {1 - \eta } \right)K_{t}$]

$$C_{t} + p_{t} H_{t} = w_{t} L_{t} + r_{t} \left( {1 - \eta } \right)K_{t} .$$

(4)

In period t + 1 parents earn income from work using the education acquired in period t, [$W_{t + 1} H_{t}$].¹⁹ These funds are used for consumption, [C_t+1], to accumulate capital to leave as a bequest, [$\left( {1 - \sigma } \right)K_{t + 1}$], and investment in the education of their offspring, [$p_{t + 1} H_{t + 1}$]

$$C_{t + 1} + \left( {1 - \sigma } \right)K_{t + 1} + p_{t + 1} H_{t + 1} = W_{t + 1} H_{t} .$$

(5)

Capital stock in period t + 1 is the capital stock left (salvaged) after government confiscation

$$K_{t + 1} = \left( {1 - \eta } \right)K_{t} .$$

(6)

What makes the above model suitable to analyze a socialist economy is the fact that taxation is almost sure to be high, or as high as possible (i.e., confiscation). This means that the values of σ, the tax on bequests, and η, the tax on capital, are close to or equal to one. These cases are analyzed ahead.

Maximizing (3) taking into account the constraints (4), (5) and (6) yields the following first order conditions

$$L_{t} :\frac{{w_{t} }}{{C_{t} }} - \lambda '\left( {L_{t} } \right) = 0,$$

(7)

$$H_{t} : - \frac{{p_{t} }}{{C_{t} }} + \delta \frac{{W_{t + 1} }}{{C_{t + 1} }} + \nu '\left( {H_{t} } \right) = 0,$$

(8)

$$H_{t + 1} : - \delta \frac{{p_{t + 1} }}{{C_{t + 1} }} + V'\left( {H_{t + 1} } \right) = 0,$$

(9)

and

$$K_{t + 1} : \frac{{r_{t} }}{{C_{t} }} - \frac{{\delta \left( {1 - \sigma } \right)}}{{C_{t + 1} }} = 0.$$

(10)

Any model of a socialist economy has to be consistent with two stylized facts: (1) the gradual erosion of the standard of living, and (2) the ever-present risk of famine. Our model yields both results. We can see this by rewriting the Euler equation (10)

$$\frac{{C_{t + 1} }}{{C_{t} }} = \frac{{\delta \left( {1 - \sigma } \right)}}{{r_{t} }} = \frac{{\left( {1 - \sigma } \right)}}{{r_{t} \left( {1 + \rho } \right)}} < 1.$$

(11)

Given that the rate of time preference ρ is positive, the above result implies that consumption at t + 1 is smaller than consumption at period t, $C_{t + 1} < C_{t} ,$ showing the gradual erosion of the standard of living. In addition, if bequests are heavily taxed/confiscated (one of the recommendations of Karl Marx and Friedrich Engels in the Communist Manifesto), which implies values of σ close to 1, (11) implies that consumption will be close to zero-there is famine. The best examples of this phenomenon are the Holodomor holocaust in Ukraine (1931–1933) and the famine in China (1958–1962) (e.g., see Nove 1992; Dikӧtter 2010; Bianco 2016).

We now wish to examine the impact of bequest confiscation on education. Substituting (8) and (9) into (10) yields

$$\frac{{\nu '\left( {H_{t} } \right)}}{{p_{t} }} = \frac{{V'\left( {H_{t + 1} } \right)}}{{p_{t + 1} }}\left( {1 - \sigma - \frac{{W_{t + 1} }}{{p_{t} }}} \right).$$

(12)

The impact of bequest confiscation is seen in (12). Increases in confiscation increase the education of parents and reduce the education of children. If the government confiscates a share of the bequests so that $\sigma \in \left( {0,1} \right),$ (12) can differentiated to evaluate the impact of the tax on bequests on parents’ education

$$\frac{{dH_{t} }}{d\sigma } = - \frac{{p_{t} V'\left( {H_{t + 1} } \right)}}{{p_{t + 1} \nu ''\left( {H_{t} } \right)}} > 0.$$

(13)

Since $\nu '' < 0$, (13) shows that a positive tax on bequests increases parents stock of human capital or education. A similar analysis holds for the education of children H_t+1, which is impacted negatively by bequest taxes. From) (12) we have

$$\frac{{dH_{t + 1} }}{d\sigma } = \frac{{V'\left( {H_{t + 1} } \right)}}{{\left( {1 - \sigma - \frac{{W_{t + 1} }}{{p_{t} }}} \right)V''\left( {H_{t + 1} } \right)}} < 0.$$

(14)

Given that $V'' < 0$, (14) shows that a positive tax on bequests reduces children education.

We now examine the impact on education of the confiscation of K_t. For simplicity, assume property is seized, that is η = 1. In this case K_t+1 = 0. Taking this into account, the budget constraints become

$$C_{t} + p_{t} H_{t} = w_{t} L_{t}$$

(15)

and

$$C_{t + 1} + p_{t + 1} H_{t + 1} = W_{t + 1} H_{t} .$$

(16)

Substituting (15) and (16) into the Euler equation, (11), and differentiating, yields the maximum positive impact of parents’ education on children education

$$\frac{{dH_{t + 1} }}{{dH_{t} }} = \frac{{W_{t + 1} + \delta \left( {1 - \sigma } \right)p_{t} }}{{p_{t + 1} }} > 0.$$

(17)

Equation (17) shows that education of parents and children are complements, and that confiscation of capital amplifies this effect. Finally, we should note that our model has implications for nonsocialist countries. One implication is that increases in the tax on bequests should decrease investments in education by both parents and children.

Appendix 2: Property loss in Eastern Europe

During WWII and the postwar period, many property-owners, businesses, and employees in Eastern Europe experienced deep disruptions to their wealth and income. Properties and businesses were physically destroyed by war, or under post-WWII communism were confiscated or collectivized with full, partial, or no compensation, or regulated such that owners lost practical control and the properties and businesses lost all their value. Many employees similarly lost their jobs or experienced punitive taxes. Effects varied across countries and regions, ethnic groups, types of property (e.g., farmland vs. apartments), and time periods. Many people, of course, in particular Jews, were killed during WWII and lost their lives along with their property.

Private property rights had been in flux throughout Eastern Europe before the advent of communist regimes in the late 1940s. Invasions, first by Germany and other countries during WWII, and then by the Soviet Union, destroyed much physical property and resulted in property confiscations and population movements. Many, whether fleeing or being expelled, “abandoned” properties, which were formally reassigned to invading governments, to private parties of the invading countries, or to local collaborators (Herman 1951).

The main target of confiscation by Germany was Jewish property. Other targets included foreign capital and large stock-owned companies such as banks. Germany also engaged in massive confiscations of all kinds of property of ethnic Poles in Western Poland, including land, farms, furniture, and tools. Confiscation of economic value, of course, was not limited to property, but could also affect labor, through punitive labor taxes, forced labor, and deportation (Wachenheim 1942; Herman 1951; Blacksell and Born 2002).

There were many differences across the countries of Eastern Europe prior to WWII. Czechoslovakia was a developed industrial democracy, whereas Bulgaria was a poor agricultural monarchical dictatorship. There were also differences under communism. Both Hungary, which collectivized farming during the communist period (Swain 1985; Berend 1990), and Poland, which largely maintained private agriculture (Dadak 2004), were somewhere between Bulgaria and Czechoslovakia in terms of economic development (Katchanovski 2000). Countries also differed considerably in the percentage of Jews and other minorities in the population and in the extent of discrimination against these groups. Also, some of these countries were part of the Allies during WWII, while others were part of the Axis (Pearcy and Dickson 1996).

While there were devastating effects throughout, the extent of property loss varied greatly. Poland and Hungary incurred more physical damage to property, while Czechoslovakia and Bulgaria incurred less (Herman 1951; Enyedi 1967). Not being invaded by Germany, some German allies, such as Hungary, experienced fewer confiscation-related upheavals in property rights during the war, with confiscations largely restricted to Jewish property (Spulber 1954).

During WWII, there were large population transfers. Transfers aimed to create more ethnically homogeneous territories and to reduce perceived “fifth columns”. Thus, German minorities left Italy. There were population transfers out of the Baltics and modern-day Moldova. Croatia expelled its Slovene population. Bulgaria and Romania forcibly swapped each other’s minority populations. Hungary and Romania permitted voluntary swaps of each other’s minority populations. The Soviet Union transferred Ukrainians out of Eastern Poland. Jewish and Roma (Gypsy) populations throughout Eastern Europe were transferred to concentration camps and murdered (Wachenheim 1942; Sieradzka and Lerman 1994; Blacksell and Born 2002).

The treatment of the properties of the “transferred” populations varied widely. In the most favorable cases, the transferred persons were allowed to take moveable personal belongings, including automobiles, livestock, professional tools, etc. However, the transferred populations were permitted to take with them only limited amounts of financial assets, such as cash, and, of course, none of their housing or land. The transferred populations were supposed to file for compensation for their foregone property with the receiving government. Foregone properties were sold by trustees or governments to pay for the damages filed by incoming (not outgoing) transferred populations, or for other purposes. Commitments to compensate transferred populations were sometimes largely honored (as in in the case of Germans leaving Italy), but often they were not, as with the population exchanges between Romania and Bulgaria (Wachenheim 1942).

In another example of the variations in population transfers and property confiscations across countries, regions, and ethnic groups, Bulgaria provides the sole example of German-allied territory in Eastern Europe where the Jewish population was not deported or exterminated (Bachvarov 1997). However, Gelber (1946) points out that Jews did suffer persecution in Bulgaria during the war, and Bowman (1986) points out that almost no Jews survived the Bulgarian occupation of northeastern Greece.

During the final stages of WWII and immediately afterward (1944–1948), many countries under the occupation of the Soviet Red Army engaged in radical land reform. While communist parties in the region planned to eventually nationalize or collectivize all farm land, the early postwar land reforms did not involve such steps. Instead, early postwar governments (not yet formally dominated by communist parties) engaged in various land policies that sought to gain the support of formerly landless peasants. Thus, these governments expelled large numbers of ethnic Germans (particularly from Western Poland, from former East Prussia, and the Sudetenland regions of Czechoslovakia that had been invaded and occupied by Germany in 1938) and confiscated their property. Those identified as traitors or collaborators also had their properties confiscated (Herman 1951; Spulber 1954; Schmitt 1993). Land confiscations originally did not affect small landowners from each country’s majority ethnic group (e.g., of Poles in Poland). However, as confiscations proceeded, new laws lowered the boundary above which land holdings were to be confiscated, falling over time and across countries from 575 hectares to 100, 50, and as low as 20 in Bulgaria. These policies, which appear to have been largely popular, reduced the political power of large landowners (Sanders 1950; Spulber 1954; Enyedi 1967).

However, governments throughout Eastern Europe did not nationalize most of the 20 million hectares of land that they confiscated, instead redistributing 12 million hectares of farm land to 65 million peasants, thus yielding large numbers of small farms (averaging one to five hectares). Beyond gaining the support of formerly-landless peasants, a key goal of these policies was to provide incentives for majority-ethnic populations to move to formerly German-held farms and lands (Sanders 1950; Spulber 1954; Enyedi 1967; Schmitt 1993). To receive land, peasants agreed to pay governments the equivalent of 1 year’s harvest, spread over 10–20 years. The proceeds were supposed to be used to compensate nationals (i.e., those not expelled) whose properties had been confiscated. Compensation payments, even though small, were made in some cases, such as Hungary, but not others, such as Poland (Sanders 1950).

While radical, early post-WWII land reforms (i.e., confiscations and redistributions) were not a completely unprecedented break in property rights in Eastern Europe. Rather, they largely continued, if more radically, long ongoing practices in the region. For instance, Austrian-owned latifundia (i.e., very large land holdings) in Czechoslovakia had been confiscated and redistributed among Slavs after World War I (Tomasevich 1958; Enyedi 1967; Schmitt 1993). Moreover, as landholding patterns throughout Eastern Europe varied widely before WWII, the extent of confiscations after WWII also varied. Before WWII, latifundia covered more than 40% of farm and forest land in Poland, Czechoslovakia, and Hungary, but less than 2% in Bulgaria. In all cases, the shares of farmland owned by non-expelled small landowners that were not confiscated were quite large: Poland, 50%; Czechoslovakia, 62%; Hungary, 62%; and, Bulgaria, 98%. The percentages of confiscated land that were redistributed (and not retained by governments) also varied: Poland, 44%; Czechoslovakia, 38%; Hungary, 58%; and, Bulgaria, 72% (Sanders 1950; Spulber 1954; Tomasevich 1958; Enyedi 1967).

A few years after the end of WWII, communist parties throughout Eastern Europe engaged in a variety of actions (largely coups) that replaced somewhat-pluralist governments with governments that were clearly communist-dominated. As the iron curtain was drawn, during 1948-1953 the new governments pushed forward with programs of forced collectivization of the remaining, already small, private farms (Tomasevich 1958; Enyedi 1967; Schmitt 1993). Collectivization, however, did not involve formal confiscation, as individual farmers retained legal title to their land. Moreover, forms of collective farms varied widely. In some cases, farms collectivized only the ownership and use of large farming equipment, implying that returns to individual farmers largely depended on the size of their land holdings and their individual effort. In other cases, farms were almost fully collectivized with returns depending only on the number of hours worked, independent of the legal title to land. In nearly all cases, however, collective farms permitted members to “retain” small plots (e.g., half a hectare) that they could farm for their own consumption (Tomasevich 1958; Enyedi 1967; Fisher and Jaffe 2000).

The degree of pressure to replicate the Soviet centralized economic model also varied across Eastern European countries. For instance, collectivization in Poland never reached 25% of farm land and was being dismantled by the late 1950s. Thus, the percentage of farm land that was operated in manners that could be recognized as privately-owned varied widely and, as late as the 1970s and 1980s, totaled 6.1% in Czechoslovakia, 10% in Bulgaria, 13.7% in Hungary, and 78% in Poland (Tomasevich 1958; Enyedi 1967; Schmitt 1993).

Confiscation of property other than rural land differed even more widely across Eastern European countries. At one extreme, Czechoslovakia formally expropriated stock-owned companies, small businesses, urban land, and apartments, but not single-story homes. In contrast Bulgaria, which expropriated small businesses, permitted families to retain ownership of one residential and one vacation home. In Hungary, housing remained largely in private hands. Nationalization and collectivization reached least deeply in Poland, covering stock-owned companies, but leaving housing and, as discussed above, also farming largely in private hands. In many cases throughout Eastern Europe, small businesses were not formally confiscated. Instead, communist governments eliminated their economic value through high taxes or regulation, such as making it difficult for private business to obtain necessary inputs from state-owned companies that were becoming more economically dominant (Fisher and Jaffe 2000).

Appendix 3: Social stratification in Eastern Europe: complete results

See Tables 13, 14, 15 and 16.

Table 13 Bulgaria

Table 14 Czechoslovakia

Table 15 Hungary

Table 16 Poland