Brian D. Silver

Who Overreports Voting

The effects of respondent characteristics with regard to the propensity of nonvoters to report that they voted are examined by analyzing the vote validation studies conducted by the University of Michigan Survey Research Center in 1964, 1976, and 1980. Previous research has suggested that vote overreporting derives from the respondents wish to appear to engage in socially desirable behavior. This earlier research suggests that the only respondent characteristic that is strongly related to overreporting is race; measures of socioeconomic status and of general political attitudes are said to be at most weakly related to the tendency to exaggerate voting. These earlier conclusions are incorrect. We measure the extent of overreporting for the population “at risk†of overreporting voting: those who did not actually vote. Respondents most inclined to overreport their voting are those who are highly educated, those most supportive of the regime norm of voting, and those to whom the norm of voting is most salient—the same characteristics that are related to the probability that a person actually votes. Blacks are only slightly more likely to overreport voting than whites. The pattern of relations between education and vote overreporting is opposite what would be found if those who falsely reported voting fit the typical image of the uneducated, uninvolved, “acquiescent†respondent who is concerned primarily with pleasing the interviewer.

Stereotype Threat and Race of Interviewer Effects in a Survey on Political Knowledge

Social desirability is generally thought to underlie the propensity for survey respondents to tailor their answers to what they think would satisfy or please the interviewer. While this may in fact be the underlying motivation, especially on attitudinal and opinion questions, social desirability does not seem to be an adequate explanation for interviewer effects on factual questions. Borrowing from the social psychology literature on stereotype threat, we test an alternative account of the race-ofinterviewer effects. Stereotype threat maintains that the pressure to disconfirm and to avoid being judged by negative and potentially degrading stereotypes interferes with the processing of information. We argue that the survey context contains many parallels to a testing environment in which stereotype threat might alter responses to factual questions. Through a series of framing experiments in a public opinion survey and the reliance on the sensitivity to the race of the interviewer, our results are consistent with expectations based on a theory of “stereotype threat.” African American respondents to a battery of questions about political knowledge get fewer answers right when interviewed by a white interviewer than when interviewed by an African American interviewer. The observed differences in performance on the political knowledge questions cannot be accounted for by differences in the educational background or gender of the respondents.

Infant mortality in the Soviet Union: regional differences and measurement issues

Changes and differences in definitions reporting procedures and completeness of count can make comparisons over time or between countries hazardous. This article illustrates the general problem by examining Soviet definitions of infant mortality and the procedures by which Soviet statistics on infant mortality are collected. The rates of infant mortality reported in Soviet official statistics substantially understate the true rates. This analysis suggests that the Soviet definition makes the reported infant mortality rates low relative to what they would be if calculated according to the definition of infant mortality by the WHO. This analysis also has implications for the analysis of Soviet mortality rates in general as well as for comparison of Soviet rates to those in other countries. Removal of a large portion of the deaths of extremely-high-risk newborns from reported infant mortality statistics affects both the level and shape of reported mortality curves. After adjusting the reported infant mortality rates for the Soviet Union and for regional subpopulations published Soviet life tables were used to assess the consistency between infant mortality rates and mortality at older ages. It was shown that the relation between the reported infant mortality rates and mortality at older ages differs greatly between regions of the Soviet Union.

Soviet Citizen Participation on the Eve of Democratization

We reassess the debate over Soviet citizen politics in the USSR during the Brezhnev era. We argue the need for a more complex model of citizen participation in the USSR before Gorbachev if we are to have an accurate baseline for evaluating changes in regime-society relations. We examine the connections between individual attitudes and individual behavior and show that political participation under the “old regime” was not nearly as one-dimensional and devoid of effect as many previous researchers (and current Soviet leaders) have described it. Many forms of political participation in the Soviet Union before Gorbachev did not fit the stereotype of a psychologically disengaged citizenry driven to participate only by coercion, a desire to conform, or a quest for particularized benefits from public officials.

The Presence of Others and Overreporting of Voting in American National Elections

This study uses the 1978 and 1980 vote validation studies conducted by the University of Michigan Survey Research Center to test the extent to which false claims about voting are affected by the presence of third parties during the interview. The presence of third parties during interviews is far more frequent than is commonly as- sumed. But the tendency of respondents to give socially approved answers is not af- fected by the presence of others during the interview. Thus, additional efforts to avoid contamination of interviews by eliminating third parties are not likely to reduce the exaggeration of self-reported vote. The analysis suggests that the declared intention to vote is a far more important factor in whether people falsely report voting than is the presence of others. Additional effort to understand the motivational basis of voting and nonvoting could help to account for variation in voting overreports.

Social Mobilization and the Russification of Soviet Nationalities.

This paper examines certain major demographic bases of ethnic identity change among the mass populations of the non-Russian nationalities of the USSR. Ethnic identity is defined here as an individuals affective attachment to certain core symbols of his nationality group: the group name and its historic language. The hypotheses tested concern the impact of social mobilization, contact with Russians, and traditional religion on ethnic identity change. The levels of Russification of 46 indigenous nationalities whose official national homelands have Autonomous Oblast status or higher are examined on the basis of 1959 Soviet census materials. By use of regression analysis it is shown that: (a) social mobilization is strongly conducive to the Russification of non-Russian nationalities residing in their official areas; (b) exposure to Russians is conducive to the Russification of both mobilized and unmobilized local populations, but the Russification effect of exposure to Russians is much smaller for the unmobilized than for the mobilized populations; (c) even where exposure to Russians is extensive and enduring, both socially mobilized and unmobilized Muslim ethnic groups are much less likely to be Russified than non-Muslims; it is proposed that a Muslim ethnic ideology mediates between the dynamic demographic influences on Russification and the actual manifestation of Russification.

The Validity of Survey Responses on Abortion: Evidence from Estonia*

This paper presents results of a validation survey of abortion conducted in Tallinn, Estonia in April and May 1992. The sample was drawn from patient records in a maternity hospital. Women who had an abortion in that hospital in 1991 were asked about recent abortions as part of a survey about women’s health. More than 80% of the respondents reported having a recent abortion. Some respondents misreported their abortion as a miscarriage. Moreover, some variation in reporting was associated with respondents’ characteristics. Ethnic Estonians were less likely to report their abortion than were Russians, women over age 40 were less likely to report the abortion than younger women, and women who had the abortion late in the first trimester were less likely to report that abortion. There was some evidence that unmarried women were less likely than married women to report their abortion, and that women who had borne three or more children were less likely to report their abortion than women who had borne fewer children. These differences probably stem from the extent to which pregnancy or abortion is considered stigmatizing for women in different situations.

Demographic sources of the changing ethnic composition of the Soviet Union

The Soviet Union is a multi-ethnic country and ethnic loyalties are an enduring aspect of Soviet society. This article discusses the changing ethnic composition of the Soviet population as a whole and by region using census data for 1959 1970 and 1989. It examines the proximate demographic sources of ethnic change for various regions of the country with special emphasis on fertility and migration. Differences in population growth rates during the last 30 years have led to marked changes in the ethnic composition of the USSR and such changes will continue in the future. The impending decline of ethnic Russians to a minority within the Soviet Union will have a greater political and symbolic significance that demographic significance. Nevertheless differential growth rates of Russians Slavs other non-Muslims and Muslims portend more dramatic changes in the ethnic composition of the USSR over the longer term. The patterns of ethnic composition of Soviet regions are not only changing but are also highly diverse. While the non-European republics are becoming more homogeneous the European republics are becoming more Russianized. In contrast to the Baltic Republics most other republics have experienced increasing ethnic indigenization for many years. The movement toward greater dispersal of governmental autonomy within the federal system accompanied by the growth of local popular movements and parties is likely to mean that the trend toward indigenization of these non-Russian republics will not be reversed in the near future.

Estimating russification of ethnic identity among non-Russians in the USSR

Many scholars have asserted that extensive ethnic russification has occurred in Tsarist Russia and the Soviet Union, but little effort has been devoted to making quantitative estimates of the extent of assimilation or of the differences among ethnic groups in tendencies to assimilate. A method of estimating ethnic reidentification in the USSR is described and evaluated. The expected number of survivors to the second census date for each age cohort of a given ethnic group is compared to the reported number of cohort members at the second census date to determine the net number who have changed their ethnic self-identification between censuses. The method establishes relative differences in the propensity to reidentify ethnically among 26 Soviet ethnic groups. Many small ethnic groups in European Russia are found to have very high estimated rates of ethnic russification; non-Russian Slavs—Ukrainians and Belorussians—are found to have fairly low estimated rates of ethnic russification. The implications of the Russian gain in population through ethnic reidentification also are assessed.

Measuring Political Culture in Multiethnic Societies Reaggregating the World Values Survey

Comparative studies of mass political culture based on surveys, such as the World Values Survey (WVS), typically leap to using aggregate-level statistics for the entire population. No previous analyses of the WVS have examined the value differences associated with a common source of cleavage: ethnicity. The authors test for ethnic differences on 10 democratic values in 16 WVS countries from 1990 to 1993. Ethnic differences within countries on these indicators are often far larger than the aggregate differences between countries. Of 259 paired comparisons between the majority and the minority groups within the 16 countries on the 10 indicators, by chance alone the authors should have found about 13 statistically significant differences; instead, they found 134. Thus, the differences in support for democratic values between ethnic groups within countries are far from just a random phenomenon.