All European countries published statistics on annual outward movement. They are summarised in Ferenczi and Willcox (1929–31, I, 236–88). These data were usually compiled from counts of passenger sailings, or from the issue of emigration passports (Finland, Italy, Hungary in some years), or from the contracts made between ships' masters and the emigrants (Denmark, Sweden). In addition, all the destination countries recorded annual inward movement. We must use these data with some caution, as can be seen when we compare departures and arrivals. Overseas arrivals appear to have exceeded European departures by some 10 million. In the early years, the recording of outward movement was inefficient in some European countries. There was also some clandestine emigration – commonly to avoid military service. One place where we would expect to find clandestine emigration was where ethnic minorities were liable to military service in an ‘alien’ army.

The greatest source of difficulty is that European countries were much less likely to record the annual number of arrivals, and destination countries were less likely to record departures. This is a problem because return migration became common. By the early twentieth century most countries had a return migration rate exceeding a quarter of outward movement and some countries had much higher return migration rates (Gould, 1979, 609). Consequently, we may not know if an increase in emigration that we are observing was partly countervailed by an increase in returns. This raises some important conceptual issues which we will discuss below.