This paper reports the results of research and analysis undertaken by Census Bureau staff. It has undergone a Census Bureau review more limited in scope than that given to official Census Bureau publications. This report is released to inform interested parties of ongoing research and to encourage discussions of work in progress.

Proceedings of the U.S. Census Bureau Conference hosted by the Council of Professional Associations on Federal Statistics.

Preface

The genesis of this conference was the Census Bureau’s release of population projections for the United States in January 2000. These showed the possible future population in great detail for each year out to 2100: (a) by single year of age to 100 plus, (b) by sex, (c) by four race groups, (d) by Hispanic and non-Hispanic, and (e) by foreign-born and native. Innovations in this product included: (a) dynamic forecasts of migration, (b) use of the Lee-Carter approach to project life expectancy, (c) incorporation of both a “true” and a “census-level” of undocumented migration, and (d) creation of projections by nativity. The publicity for this document centered around the finding that the U.S. population would double during the new century, reaching almost 600 million in size. There was very little feedback from any group about the results or the methodology.

However, some experts did express concern about the fertility assumptions and methodology. Their primary issue was with the assumption that future fertility would remain somewhat higher than it is now. The overall total fertility rate was projected to rise from 2048 in 1999 to 2207 in 2025 and stay near that level through 2100. The essential reason for this projected increase was our decision to adopt birth expectations data as the primary predictor of future fertility. Moreover, we chose to assume that the “ultimate fertility level” of each race-ethnic group in 2150 would still be at 2100 (the so-called “replacement level” of fertility)—a higher fertility level than the U.S. total population had actually experienced since the early 1970’s. This assumption was made based on our application of current forecasting models to the most recent birth expectations data plus the trend oddities, which we found to occur if we attempted to converge a number of race/ethnic groups to any lower fertility level or converge their fertility any more rapidly. Finally, the above-replacement fertility assumption also received considerable attention because the United Nations had just released projections for each of the world’s developed nations in which virtually every developed nation was assumed to always have a total fertility rate under 1800.

Because of these general concerns that our fertility assumptions for America were too high, and because we could find no methodology that would yield significantly different results, we thought this was an excellent time to host a conference devoted to this topic. As summarized in the invitation letter, the charge to the conference attendees was the following:

“While there have been dramatic changes in fertility in the rest of the world, American fertility has remained remarkably constant since the early 1970’s. Our current projections mirror this stability and are inconsistent with the lower assumptions used in other projections. From this meeting we especially hope to learn about the contribution of compositional aspects such as race, ethnicity, and nativity. We also wish to better understand the nature of this apparently anomalous stability in American fertility and to monitor any nascent signs of pending changes in fertility in the United States.”

Three separate papers were commissioned and two discussants were assigned to provide formal comments. Every attendee received the commissioned papers several weeks in advance of the conference. A professional scribe was hired to take down all the comments of other participants.

This report contains all of this information and attempts to accurately convey the opinions of all.

The commissioned papers and formal discussions are shown exactly as they were presented. The comments from the audience were summarized by Gregory Spencer from the scribe’s notes. He apologizes for any misstatements or misinterpretations ascribed to the participants.