Overt terrorist acts and the specter of terrorism impose significant costs on the community of civilized nations. These costs can be reduced by taking retaliatory action against terrorist organizations and the countries that sponsor them. Assuming a positive range over which retaliatory cost is less at the marginal than the resulting benefit from reduced terrorism, there exists some positive level of retaliation which is efficient from the perspective of the victimized countries. An obstacle to achieving the efficient level of retaliation against terrorists results from the fact that much of the benefit from retaliation is general and cannot be captured entirely by the retaliating country. A fully efficient policy of retaliation will therefore require a cooperative response from all victimized countries. Achieving this cooperation, however, may confront the well-known free rider or prisoner's dilemma problem. While collectively the targets, and potential targets, of terrorism are better off retaliating, each will likely see its advantage best served by not retaliating. In this case it is quite possible that no one will retaliate when the best response is for everyone to retaliate. A closer look at the problem of terrorism, however, suggests that the standard prisoner's dilemma may not be a significant obstacle to retaliation. Under plausible conditions, it will pay one country to retaliate even if other countries choose to free ride. Indeed, given the two options in the standard prisoner's dilemma setting, (cooperating (retaliating) or noncooperating (nonretaliating), it is very likely that all victimized countries will retaliate; that no country will take a free ride. But this does not mean that the prospects for genuine cooperation in confronting terrorism are good. In the case of retaliation against terrorist, free riding is only one way a country can increase the net benefit it receives from the public good provided by another. It is possible for one country to, in effect, sell the public good of reduced terrorism that is generated by the retaliation of another country. This paidrider possibility has important, and obviously adverse, implications for the prospects that retaliation will result from a cooperative effort on the part of those countries victimized by terrorists. In Section I, the implication of free riding and paid riding for retaliation against terrorists will be investigated in a two-country setting. In Section II, the circumstances that either encourage or discourage a country from being a paid rider on the retaliation of another country will be considered and some real world examples of paid riding will be discussed.

Viewing the United States as comprising many racial and ethnic groups, it is shown that group differences in earnings, schooling, and rates of return from schooling are striking and that the groups with higher levels of schooling also have higher rates of return. These data are shown to be consistent with a child quality investment model, but they are not consistent with the hypotheses that the primary determinants of schooling level are discrimination, minority group status, differences in time preference (discount rates), or "tastes" for schooling. Group differences in fertility and female labor supply are examined as partial determinants of investment in child quality. Policy implications are discussed.

This article develops a theory in which concentrated-trading patterns arise endogenously as a result of the strategic behavior of liquidity traders and informed traders. Our results provide a partial explanation for some of the recent empirical findings concerning the patterns of volume and price variability in intraday transaction data.