First, many other factors affect election results, and they vary geographically, potentially producing patterns similar to those related to fraud. Concentrations of votes in small regions, in conjunction with high turnout, may result from the territorial concentration of socio-economic or cultural minorities, the prominence and mobilising power of local candidates, or small communities (Coleman, 2004). Even in old democracies, turnout tends to correlate with partisan vote (Grofman et al., 1999). Uneven distributions of digits can result from the design of electoral districts, candidate nomination strategies, or altering voting strategies across electoral constituencies (Cain, 1978; Bochsler, forthcoming; Mebane, 2013). Also, “innocent errors” by election administrators or poll workers may be indistinguishable from fraud by statistical methods.