ADDITIONAL EVIDENCE ON THE SOCIAL CHARACTERISTICS OF RIOT CITIES.

In his conclusion, Downes states that "cities experiencing hostile out- bursts have a very distinctive set of social, economic, and even political structural characteristics."19 It is true that Downes expresses reservations about his analysis, cautioning the reader that"… this analysis represents a preliminary examination of the information we have collected,"20 and adds that he intends to undertake more extensive analysis of his data.21 By coincidence, our work involves more sophisticated analysis of data related to some of the variables which Downes studied. Furthermore, while Downes treats variables which characterize city populations in toto, and deduces from these the conditions of nonwhite persons in the cities, our study employs direct measures of characteristics of the nonwhite population and of differentials between the white and nonwhite groups. Of course, our study examines just a few socioeconomic variables, using only one bit of demographic data, and is thus less broad in scope than Downes'. Nevertheless, the results using these variables cast doubt on some of Downes' conclusions. Thus, contrary to Downes' conclusion, we find that the per cent of the population in SMSA's that is Negro does not appear to be related to the occurrence or intensity of civil disturbances. Where Downes finds low educational levels to be related to outbreaks of violence, our results are equivocal on this score. In the study most closely related to Downes (Study I), we are in partial agreement on the role of education; however, when an intensity index is used or where white-nonwhite differentials are studied, no clear relationship is observed, Furthermore, where Downes concludes from per capita income data for entire city populations that income levels are inversely related to occurrence and intensity of disturbances, we find almost the reverse to be true when nonwhite income levels are studied. When white-nonwhite income differentials are used, the results are again equivocal. Using crowding as a surrogate for housing quality we find no relationship between housing quality and the disturbances, where Downes concludes that this characteristic may be important. On the other hand, we find that home ownership, which may be a kind of proxy for community stability, is related to the disturbances in both their existence and severity. In fact, in this case the white-nonwhite differential becomes important. Finally, it is important to emphasize that our results do not go very far in "explaining" the causes of the disturbances. As Cohen has said, "One of the major errors in both the white and Negro communities has been seeking a single answer to a complex of problems. Integration, for example, has now given way to a theory of upgrading the slums. Status frustration has `lost ground to economic frustration. These are false dichotomies. The racial problem is a complex of many problems which have accumulated over the years."22 Relatively simple deprivation models of the type discussed in Downes' paper may shed light on some of the roots of urban discontent, but are over simplifications. In a way, their failure indicates the existing lack of understanding of these complex and urgent problems. [ABSTRACT FROM AUTHOR]