Abstract:The eastern coastal region

We argue that economic inequality between ethnic groups increases state repression. We contend that a high level of ethnic inequality fuels distributional conflicts between poor and rich ethnic groups. It also increases the salience of ethnic identity and promotes ethnic mobilization to challenge the status quo. This between-group tension creates collective grievances for ethnic groups, mounts challenges to incumbent governments and increases perceived threats to governments. The greater the perceived threats, the more likely that governments will employ coercive measures. We further argue that the impact of ethnic inequality on state repression is moderated by the level of democracy. Various institutional mechanisms in democracies increase the costs of repression, reducing leaders’ incentives to employ coercive measures, even when facing high levels of ethnic inequality. Evidence from 152 countries between 1992 and 2011 supports our arguments.

No abstract

This paper investigates the domestic political factors that shape the participation of autocratic regimes in bilateral investment treaties (BITs). We argue that autocratic time horizon positively affects governments’ motive to sign BITs by influencing the costs of complying with investor protection standards included in the treaties. These treaty provisions severely constrain discretionary policy maneuvers that are critical to autocratic survival. Autocratic regimes expecting to rule for a considerable time period are willing to relinquish some discretionary policy space in the interest of enhancing the credibility of their investor protection commitment—and hence promoting investment inflows. However, autocratic governments with short time horizons rely heavily on discretionary policy maneuvers to stabilize their grip on power and are likely to infringe on investors’ interests to extract resources to ensure their political survival, making the costs of compliance with BITs too high to bear. Using a country-dyad data set of BIT signatures from 1971 to 2009, we find strong support for our argument.

No abstract