for useful comments and suggestions. All remaining errors are our own. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research. NBER working papers are circulated for discussion and comment purposes. They have not been peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies official NBER publications.

Yale for many helpful suggestions. We are grateful to Mike Chernew for the access to, and Tami Swenson for clarification of, the Medicare Current Beneficiary Survey data. All remaining errors are our own. Fang and Silverman gratefully acknowledge financial support from the Economic Research Initiative on the Uninsured. The views expressed herein are those of the author(s) and do not necessarily reflect the views of the National Bureau of Economic Research.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.We propose entertainment and travel costs (ETC) expenditures as a measure of corruption in Chinese firms. These expenses are publicly reported in firms' accounting books, and on average they amount to about 3 percent of a firm's total value added. We find that ETC is a mix that includes grease money to obtain better government services, protection money to lower tax rates, managerial excesses, and normal business expenditures to build relational capital with suppliers and clients. Entertainment and travel costs overall have a significantly negative effect on firm productivity, but we also find that some components of ETC have substantial positive returns to firms.

Black motorists in the United States are much more likely than white motorists to be searched by highway troopers. Several recent lawsuits against state governments have used this racial disparity in treatment as evidence of "racial profiling," a term that refers to the police practice of using a motorist's race as one of the criteria in their motor vehicle search decisions. Racial profiling originated with the attempt to interdict the flow of drugs from Miami up Interstate 95 to the cities of the Northeast. For example, in 1985 the Florida Department of Highway Safety and Motor Vehicles issued guidelines for police on "The Common Characteristics of Drug Couriers," in which race/ethnicity was explicitly mentioned as one characteristic (Robin S. Engel et al., 2002). While the initial motivation for such guidelines may have been to increase the troopers' effectiveness in interdicting drugs, it also unfortunately opened up the possibility for troopers to engage in racist practices against minority motorists.Following the public backlash generated by several cases in the 1990s such as Wilkins v. Maryland State Police (1996) and Chavez v. Illinois State Police (1999), almost all highway patrol departments have denounced using race as a criterion in stop-and-search decisions. But many citizens, especially minorities, are skeptical of this claim: motor vehicle search decisions, by their very nature, are made in the midst of face-to-face interactions, and thus it is simply hard to imagine that troopers can block the race and ethnicity information that a motorist presents. Moreover, data on trooper searches continue to show that troopers tend to search a higher proportion of minority motorists than white motorists. As is now well known, however, racial disparities in the aggregate rates of stops and searches do not necessarily imply racial prejudice (see, for example, John Knowles et al., 2001;Engel et al., 2002). If, for example, black drivers are more likely than white drivers to carry contraband, then the aggregate rate of stops and searches would be higher for black drivers even when race was not a factor in troopers' decision-making. Moreover, racial profiling may also arise if police attempt to maximize successful searches, and race helps predict whether a driver carries contraband. This situation is called statistical dis-