8. Until the same date an alien arrival in transit to some other country was deemed an immigrant, but since that date such persons have been classed as non-immigrant aliens. This change also makes the figures before 1903 not strictly comparable with later ones. About three per cent. of those who were formerly classed as aliens have been excluded since 1903. The alteration has brought the two definitions closer together, but in so doing has entailed administrative difficulties which lead the bureau to favour a return to the former system or at least to favour collecting the head tax from such aliens in transit.
9. An immigrant in the statistical sense is a person liable for and paying the head tax. But to this there are two slight exceptions. Deserting alien seamen not apprehended are liable for the head tax which is paid by the company from which they desert, but such cases are not included in the statistics. Citizens of British North America, Cuba and Mexico coming from other ports than those of their own country are reported as immigrants, but do not pay the head tax. Obviously both are minor exceptions hardly affecting the rule. In the popular or theoretical meaning of immigrant this head tax is not an element.
10. Probably other changes of definition have occurred of recent years. No attempt has been made to exhaust the list. The general tendency of the changes has clearly been towards a closer agreement of the popular and the statistical meanings. But they have probably tended to make the increase of immigration indicated by the figures greater than the actual increase, and to that degree to make the figures misleading. If the Government Bureau of Immigration and Naturalization could make a carefully studied estimate of the extent to which such changes in the official reports really modify the apparent meaning of the published figures, it would render a valuable service.
11. A committee like the present can hardly make such an estimate or go further than to point out that for the reasons indicated the official statistics of immigration are likely to be seriously misinterpreted and are constantly misinterpreted by the public.
The official statistics of immigration being subject to all the qualifications indicated and reflecting so imperfectly the amount of immigration as ordinarily or popularly conceived the question at once arises, Can any substitute or any alternative be proposed? What the public is mainly interested in, I think, and what it commonly but erroneously believes is indicated by the official figures of immigration, is the net addition to the population year by year as a result of the currents of travel between the United States and other countries.
Alternative figures for the last eight years, a period which closely coincides with the last great wave of immigration now at or near its crest, may be had by comparing the total arrivals and departures in the effort to get the net gain. The results appear in the following table:
| Fiscal Year | Total Passengers Arrivals | Total Passengers Departed | Total Immigration | Arrivals Minus Departure | Per Cent. That Net Increase Makes of Immigration |
| 1898 | 343,963 | 225,411 | 229,299 | 118,552 | 51.8 |
| 1899 | 429,796 | 256,008 | 311,715 | 173,788 | 55.8 |
| 1900 | 594,478 | 293,404 | 448,572 | 301,074 | 67.0 |
| 1901 | 675,025 | 306,724 | 487,918 | 368,304 | 75.5 |
| 1902 | 820,893 | 326,760 | 648,743 | 494,133 | 76.3 |
| 1903 | 1,025,834 | 375,261 | 857,046 | 650,573 | 75.9 |
| 1904 | 988,688 | 508,204 | 812,870 | 480,484 | 59.3 |
| 1905 | 1,234,615 | 536,151 | 1,026,499 | 698,464 | 68.1 |
| 1898-1905 | 4,822,662 | 3,285,372 | 68.1 | ||
The figures indicate that the net increase of population by immigration during the last eight years has been slightly more than two-thirds of the reported immigration. But these figures of net increase should be increased by an estimate of the arrivals by land from Canada and Mexico. As the Canadians and Mexicans by birth residing in the United States in 1900 were 14.2 per cent. of all residents born in other foreign countries, this would indicate an influx of 466,000 Canadians and Mexicans, a figure probably in excess of the truth since the currents have probably been setting Canadaward of recent years. I estimate, therefore, that the net increase from immigration 1898-1905 has been about 3,750,000 instead of 4,820,000 as might be inferred from the reports of the bureau of immigration. The actual increase would then be about seventy-eight per cent. of the apparent increase.