personal injury, enforcing community moral standards) that generally give priority to violent offenses. Moreover, although it is true that violent crimes are not uniformly feared more than other offenses, it would be a serious mistake to ignore such crimes because of the enormous fear they are capable of producing. The last column in Table 1 shows the expected fear score for each of the offenses under the multiplicative model, with perceived risk set to an arbitrary constant (i.e., 5). If all crimes were perceived to be equally likely, as in this example, violent crimes would clearly outweigh all other forms of crime in the fear they evoke. That is an unlikely scenario, to be sure, but the point is that even moderate increases in the perceived risk of violent victimization have the potential to increase fear enormously.

One of the most distinctive features of fear of victimization is that fear, like victimization itself, is not randomly distributed in the population. Evidence accumulated over the past two decades consistently indicates that fear is particularly pronounced in two groups: females and older individuals (Hindelang et al., 1978; Warr, 1984; Skogan and Maxfield, 1981; Baumer, 1978; Clemente and Kleiman, 1977; DuBow et al., 1979). In their three-city survey, for example, Skogan and Maxfield (1981) found that the proportion of respondents who felt "very unsafe" walking alone in their neighborhood at night rose from 7 percent among those aged 18-20 to 41 percent among those over 60, and although 6 percent of males reported such fear, the figure increased to 23 percent among females. Hindelang et al. (1978) report much the same results, but they also note that the association between fear and age is much stronger among males than among females. These patterns are quite evident in the GSS data. The sex difference in responses to the fear item is very large, with 22 percent of males and 60 percent of females responding yes in the cumulative (1972-1987) file. Among females, this proportion is rather constant across age groups, varying no more than 6 percent. Among males, however, the age gradient is much more marked, increasing from 14 percent among those under 20 to 32 percent among those over 60.

How can such large sex and age differentials in fear be explained? One possible explanation is that females and the elderly are more afraid than others because they face the greatest objective risk of victimization. In fact, however, exactly the opposite