This is the fifth in this year's series of posts by PhD students on the job market.
How should democratic governments interact with their authoritarian counterparts? The options include initiating a trade war or facilitating access to foreign media. Throughout history, a number of democratic governments have focused on engagement policies, specifically on promoting more interactions between citizens that live in democratic and authoritarian societies. However, the effects of such policies are largely unexplored: It is unclear whether attitudes of individuals living in non-democratic societies change when they meet with individuals that are socialized in democratic societies. Moreover, it is unclear whether these engagement policies strengthen the support for democracy during democratic transitions. These questions are important: a recent theoretical literature in political economy suggests that the degree of support for democracy within a society is critical in determining whether countries transition to democracy or experience autocratic reversals (Besley and Persson, 2018).
Natural Experiment & Empirical Strategy
In my job market paper, I study a policy that was implemented during the Communist dictatorship in East Germany to address these questions. In 1972, the East German regime agreed to reduce restrictions for private visits, specifically for West Germans travelling to East Germany to visit family and friends.
To identify the effects of this policy, I leverage variation in the level of effective travel restrictions across East German districts. In particular, a subset of districts located close to the border between East and West Germany were granted simplified procedures regarding visits from West Germany. I leverage this cross-sectional variation in a spatial regression discontinuity design: My empirical strategy compares outcomes across East German districts that were located just next to each other and, as I document empirically, looked similar in many aspects prior to the introduction of this policy. However, the districts differed in one important dimension: treated districts just fell within the area for which visits from West Germany were simplified, while the remaining districts were located just outside of it and therefore were not granted such simplifications. The red line in Figure 1 highlights this discontinuity within East Germany.
Figure 1: Extended Visitors Program Zone and Policy Discontinuity
In addition to the spatial regression discontinuity design, I implement a randomization inference approach. This approach provides a sharp test of the null hypothesis that the visitors program treatment had no effect. In particular, I implement this approach by comparing the impact of the actual visitors program to the effect of counterfactual assignments of districts to the visitors program. Importantly, the counterfactual spatial treatment assignments all respect the fact that the actual set of districts participating in the visitors program formed a connected and contiguous area (as shown in Figure 1). I re-estimate the empirical specification for 500 counterfactual assignments that were randomly selected among the set of feasible treatment assignments. I then use the position of the actual treatment effect in the distribution of counterfactual treatment effects to show that the observed empirical patterns described below would have been unlikely to arise in the absence of the actual policy (see Figure 2; true effects are depicted by red vertical bars and the limits of 95% confidence intervals are represented by gray vertical bars).
Visitors program weakened the support for the East German regime
First, I show that the program was effective in increasing the number of visits, and that it substantially increased the likelihood that an East German citizen would interact with visitors from the West. Given that the policy was implemented in 1972 and the East German revolution started in the fall of 1989, a simple back-of-the-envelope calculation based on my estimates implies that, for a period of 17 years, East Germans in the treated districts had an 11% higher chance to personally interact with West Germans throughout any given year.
Second, I study the effect on support for the Communist regime using different measures of political behavior. I estimate that, on average, the assignment to the program increased the number of protest days during the fall of 1989 by 25%. These protests were critical for overthrowing the East German regime and setting of the transition to democracy (Dale, 2005). I also find that assignment to the program lowered electoral support for the legal successor of the Communist party during the democratic transition by nearly 10%. This effect persisted for almost a decade after the East German revolution.
Figure 2: Actual Treatment Effects Compared to Distribution of Counterfactual Effects
Contact weakens the pervasiveness of the East German ideology and nurtures democratic preferences
I argue that a plausible explanation for these results is that increased interactions between citizens that live in different political regimes affect attitudes and, in particular, weaken the pervasiveness of the ideology that underlies the support for the East German regime.
During these visits, East and West Germans were able to exchange views on the different political systems, differences in societal norms as well as on standards of living in East and West Germany. Direct exposure to West Germans' points of view could have been instrumental in shaping individuals' attitudes towards the dictatorship and East German society in general.
To investigate the effects on attitudes, I use individual-level survey data from the German Socio-Economic Panel measured at the time of the democratic transition. In particular, I focus on attitudes towards behavior that was specially emphasized by the East German regime: being dutiful and showing high performance at one's workplace. Historians provide detailed discussions of situations in which the East German state regularly mandated individuals to be obedient, while framing compliance as dutiful behavior. To further document the East German regime's emphasis on the desirability of such behavior, I also perform text analysis using articles published in the regime's official party newspaper.
I estimate a significant, negative effect of the program on the approval of such forms of behavior. This indicates that the policy weakened the acceptance of the ideology that was promoted by the East German regime. Importantly, I also show that the policy reduced satisfaction with democracy in East Germany by about 19 percent. These effects are stronger for cohorts that were more likely to have social ties to West Germany.
I interpret these findings in the following way: Stronger democratic values, particularly with the East German ideology being less pervasive, may lead to stronger opposition and therefore likely explain the emergence of protests against the regime during the democratic transition.
Lastly, I also provide evidence that alternative channels such as economic expectations, grievances
about repression and surveillance or the desire for national unity are unlikely to explain the differences in outcomes during democratic transition.
My findings highlight how interactions between citizens that live – and are largely socialized – in different political systems shape political beliefs and aggregate outcomes such as protest. Moreover, my empirical results imply that influencing those beliefs within authoritarian regimes matters for outcomes during democratic transitions. In addition, these results help us better understand how and why authoritarian policies of isolation work: by imposing travel restrictions, authoritarian governments seem to be more successful at preserving the pervasiveness of the ideology that contributes to popular support for these regimes.
I show that a foreign intervention that simplifies interactions between people living in different types of political systems can effectively undermine support for authoritarian rule within non-democratic societies. Examples of policy initiatives of this type include South Korea's sunshine policy towards North Korea or the easing of US-Cuba relations under the Obama administration. The evidence discussed in my paper suggests that these policies are a powerful way to change attitudes and affect behavior within non-democratic societies.
Andreas Stegmann is a PhD Candidate in Economics at CEMFI.