DeWaard, Jack, Keuntae Kim, and James Raymer. "Migration Systems in Europe: evidence from harmonized flow data". Demography, Vol.49, No.4 (2012): 1307-1333.

DeWaard, Jack, Keuntae Kim, and James Raymer. "Migration Systems in Europe: evidence from harmonized flow data". Demography, Vol.49, No.4 (2012): 1307-1333.

• Migration systems theory attempts to predict and explain migration flows through examining the historical and economic connections between sending and receiving countries (1307, 1309).
• Issues with the measurement of migration, due to both incomplete data and different national methods of data collection make it difficult to actually analyse migration using the available data, making data-dependent theories difficult to actually apply to real-world migration (1308).
• The authors attempt to solve this methodological problem in theorizing migration flows by standardizing the available data through estimations. The exact method is complicated and math-intensive, essentially calculating the differential between the receiving reports of the Nordic countries and the sending reports of all others, and assuming the overreporting or underreporting in these flows in reflected in that country's reports for all migration flows (1308-1309, 1311-1317).
• Migration between European states generally falls into three clusters of high sending and receiving activity, indicating frequent two-way migration between these states. Cluster A is composed of France, Germany, Italy, Spain, and the UK; Cluster B contains Bulgaria, Cyprus, Czechia, Estonia, Greece, Hungary, Latvia, Lithuanian, Luxembourg, Malta, Poland, Romania, Slovakia, and Slovenia; Cluster C consists of Austria, Belgium, Netherlands, Denmark, Sweden, Finland, Ireland, and Norway (1322).
• While the countries in Cluster A experience considerable migratory flows, the countries of Cluster B are much smaller levels of migration. These levels are, however, consistent across years and reflect two-way migration, as opposed to the fluctuating and one-way migration from Cluster B to Western European states (1322).
• The author conduct statistical tests of the effects of a number of factors on determining the level of migration between European states. These factors are: shared colonial or imperial past, population size, urban population, old-age dependency ratio, yearly change in labour force, geographic distance,  and social expenditure per household (1324).
• The statistical analysis revealed that larger populations encourage more immigrants, as do larger urban populations. Countries which are geographically distant have a smaller migratory flow. Economic growth promotes immigration, whereas unemployment can increase out-migration but only if people have enough money to immigrate. Shared imperial history or common language increases migration (1325-1327). Countries which become EU members during a similar time are less likely to experience cross-migration (1328).