In an article newly published in the Journal of the European Economic Association ( just in time for the South Sudan referendum!), Alberto Alesina, Janina Matuszeski and I look at the general problem of "artificial states." (Ungated working paper here.) We have one conventional and one unconventional definition of artificial states, both of them continuous measures of "artificiality." The conventional one measures the frequency of ethnic groups split in two by a border (usually one that colonizers had mindlessly created). The unconventional one measures the "squiggliness" of country borders, on the theory that colonizers drawing artificial borders were prone to drawing straight lines (see Sudan in picture), while" natural" states rarely had straight borders (see France).
We identified countries that were "most artificial" on both measures:
Chad, Ecuador, Equatorial Guinea, Eritrea, Guatemala, Jordan, Mali, Morocco, Namibia, Niger, Pakistan,Sudan, and Zimbabwe.
We also described some illustrative country cases:
Pakistan wound up as a collection of Balochistan, NWFP, Sindh (all of whom entertained secession at various times), East Bengal (which successfully seceded in 1971 to become Bangladesh, although only after a genocidal repression by West Pakistani troops), Mohajir migrants from India (many of whom regretted the whole thing), and West Punjab (which had its own micro-secessionist movement by the Seraiki linguistic minority).
Both measures predict that more artificial states are prone to worse development outcomes than less artifical ones, although the conventional measure is much more statistically robust as a development determinant than the "squiggliness" measure.
We don't draw any policy conclusions in the paper, nor will I do so in this blog post... but you can if you want.