Win,lose, or withdraw: A categorical analysis of career patterns in the House of Representatives, 1948–1978

The literature on incumbency advantage has focused on margin as an indicator for electoral security. But while electoral margin is a goodex ante measure, it is a poorex post measure of security. Further, existing work has not integrated the choice of retirement with changes in the level of security. To improve the specification and definition of “marginality”, a multinomial LOGIT model is proposed where the dependent variable is categorical. Account is therefore taken of all the ways a Representative's term in office can end, including reelection, defeat, retirement, or pursuit of other office. The sample includes all U.S. Representatives elected for the first time between 1948 and 1978. The results indicate that (1) margin in the previous election is a significantex ante proxy for the probability of electoral defeat, and (2) while the electoral safety of all incumbents in the House has been increased, the increases are greater for members elected for the first time in the period since 1965.     

Preference modification vs. incentive manipulation as tools of terrorist recruitment: The role of culture

Terrorism is a tactic much more likely to be used when combatants have asymmetric numerical strength and weaponry. Only if one side is comparatively very weak will it use terror tactics. This weakness requires a means of controlling strong incentives for free-riding or defection from the weaker side. There are two (nonexclusive) answers: (1) Atttract or inculcate recruits with an innate preference for cooperation, even if it results in the recruit's own death (2) Create a set of incentives that reward loyalty, by giving access to excludable near-public (“club”) goods. Culture is the key to achieving either of these solutions. Culture is defined here as the set of “inherited” beliefs, attitudes, and moral strictures that a people use to distinguish outsiders, to understand themselves and to communicate with each other. The primary question is whether culture creates a preference for cooperation as a primitive, or accommodates incentives such as excludable club goods that can only be obtained by cooperation. The difference between the two accounts matters greatly for determining the correct strategy to fight terrorism. If terrorists are selected for having unusual (cooperative, from the perspective of the terror group) preferences, then recruitment must be disrupted somehow. If, on the other hand, terrorists allow themselves to be recruited to gain access to club goods, then the intervention strategy must be the disruption of social networks that credibly guarantee access to those club goods.     

The prohibition-repeal amendments: A natural experiment in interest group influence

The pattern of state support for Prohibition (18th Amendment, 1919) and Repeal (21st Amendment, 1933) is analyzed and compared. This comparison is important because Prohibition is the only amendment ever to be repealed. The main thesis is that there was no wholesale change in preferences of citizens. Instead, producer interests failed to mobilize effectively in 1919, and the coupling of moral and economic arguments that worked in 1919 broke apart in 1933. Regression analysis is conducted on state legislatures (for Prohibition) and state referenda on convention delegates (for Repeal), so states are observations in the cross-sectional regression analysis. The results broadly support the main thesis.