Before addressing directly the reply of Taeuber and James, it is useful to comment on the general issue of the private sector's contribution to segregation. It is quite natural to assume that the existence of a private sector in education does contribute to racial segregation, and our finding that it does not do so may be counterintuitive. Indeed, many accept the existence of the private sector with some misgivings because it contributes to racial segregation, but seeing it as necessary because to outlaw private education would constitute an undesirable degree of coercive power of the state. Yet our analysis indicates that the misgivings are misplaced (except, as we noted in our original analysis, in certain locations, illustrated in our regional analysis by the South [Coleman, Hoffer and Kilgore, 1982a:218, which we refer to as BB]). Where is it that our analysis and intuition part company, and what accounts for the divergence? The key to the puzzle lies not in the private sector, but in the public. It lies in the opportunity for residential mobility within the public sector, together with variations in racial composition of different school districts. American education in metropolitan areas over the past few decades has been characterized by a movement out from central city schools. The inner-city schools have been seen as less desirable than others by both whites and blacks, and both have exercised choice to have their children attend other schools. In this exercise of choice, there have been two potential destinations: suburban schools and private schools. Quite obviously, the extreme growth of suburban schools since World War II, together with the approximate constant proportion of students in the private sector (at about 10 percent). means that the orinciple destination has been suburban schools. There is one well-known exception: the formation of private schools in certain parts of the South when desegregation was mandated, and both because districts are usually county-wide in the South and because blacks were not concentrated in central cities, there was no predominantly-white suburban district to move to. Even with these schools, about the same proportion of children go to private schools now as in 1954, the year of the Brown decision. The possibilities of these two modes of escape from undesirable inner-city schools have been differentially available to white families and black families. A private school required money-a rather large amount for elite private schools, less for religiously-operated schools, particularly Catholic schools. A move to the suburbs also required money, to be able to buy a house or rent in a suburban area. As is well known, housing discrimination has been the form of racial discrimination most resistant to legal measures, and the last to remain widespread. Taeuber has, in fact, been perhaps foremost among the demographers who have documented this. The result is that use of the public-sector path to escape undesirable inner-city public schools has been, until very recently, available only to whites, and still remains far more available to whites with a given income than to blacks with the same income. Use of the private-sector path has also been more available to whites because their income is on average higher; but because it is based on income rather than on income and residence, the private-sector path has been relatively more available to blacks than has the public-sector path. This fact is evident in the large number of blacks enrolled in Catholic schools, which are often geographically accessible to blacks, and which do not, as do suburban schools, limit attendance to those living within communities that have been residentially unavailable to blacks. It is also evident in our analysis (BB, Figure 3-2), which shows that the increment in probability of attending a Catholic school with an increment in income is higher for blacks than for whites at middle and upper incomes, and (as BB, Figures