There has been long and wide-ranging debate in the social science literature about how best to conceptualise and to measure segregation (see, inter alia, Allen and Vignoles, 2007; Johnston and Jones, 2010; Harris, 2011). A popular measure is the dissimilarity index, usually attributed to Duncan and Duncan (1955). This is somewhat ironic because in another paper published in the same year, the same two authors were much more cautious about advocating any one index as preferable to others and were wise to the geographical limitations: "all of the segregation indexes have in common the assumption that segregation can be measured without regard to the spatial patterns of white and nonwhite residence in a city" (p.215). Whilst one response to this shortcoming has been the development of spatial measures of segregation (Wong, 1993; Reardon and O'Sullivan, 2004; Harris, 2012), a number of papers from the 1980s and 90s treated the measurement of segregation as a (spatial) optimisation problem (Jakubs 1981; Morgan 1983; Waldorf 1993). In this paper I revisit that optimisation literature, substituting geographical distances between places with ‘nearest-neighbour distances’ to determine the cost function. Applying this method to the 2011 Census data and to England, I consider claims of “white flight” that have appeared in the media.