This edition of Projections of Education Statistics provides projections for key education statistics, including enrollment, graduates, teachers, and expenditures in elementary and secondary public and private schools, as well as enrollment and degrees conferred at postsecondary degree-granting institutions. Included are national data on enrollment and graduates for the past 15 years and projections to the year 2020. Also included are state-level data on enrollment in public elementary and secondary schools and public high schools from 2002, with projections to 2020. This report is organized by the level of schooling with sections 1, 2, 3, and 4 covering aspects of elementary and secondary education and sections 5 and 6 covering aspects of postsecondary education.
There are a number of limitations in projecting some statistics. First, state-level data on enrollment and graduates in private elementary and secondary schools and on enrollment and degrees conferred in postsecondary degree-granting institutions are not included. Further research and model development are needed to develop reliable projections of private school enrollment and graduates by state. Neither the actual numbers nor the projections of public and private elementary and secondary school enrollment include homeschooled students because more data are required to develop reliable projections.
Similar methodologies were used to obtain a uniform set of projections for each of the 50 states and the District of Columbia. These projections are further adjusted to agree with the national projections of public elementary and secondary school enrollment and public high school graduates contained in this report.
There were changes in how the exponential smoothing forecasting technique was used in this edition of Projections of Education Statistics, compared to how it had been used in previous editions. Exponential smoothing is used in the projections for elementary and secondary enrollments, high school graduates, new teacher hires, postsecondary degree-granting enrollments, and degrees conferred. In prior editions of Projections of Education Statistics, the smoothing formula used fixed coefficients and the new methodology uses variable coefficients. There has been no systematic examination to determine whether the projections computed using the new method are comparable to those using the prior method. For information about exponential smoothing used in Projections, see appendix A.0.
The summary of projections provides highlights of the national and state data, while the reference tables and figures present more detail. All calculations within Projections of Education Statistics are based on unrounded estimates. Therefore, the reader may find that a calculation, such as a difference or percentage change, cited in the text or figure may not be identical to the calculation obtained by using the rounded values shown in the accompanying tables.
Appendix A describes the methodology and assumptions used to develop the projections; appendix B presents supplementary tables; appendix C describes data sources; appendix D is a list of abbreviations; appendix E presents a list of the references; and appendix F is a glossary of terms.
Projections of a time series usually differ from the final reported data due to errors from many sources, such as the properties of the projection methodologies, which depend on the validity of many assumptions.
The mean absolute percentage error is one way to express the forecast accuracy of past projections. This measure expresses the average of the absolute values of errors in percentage terms, where errors are the differences between past projections and actual data. For example, based on past editions of Projections of Education Statistics, the mean absolute percentage errors of public school enrollment in grades prekindergarten–12 for lead times of 1, 2, 5, and 10 years were 0.3, 0.6, 1.3, and 2.4 percent, respectively. In contrast, mean absolute percentage errors of private school enrollment in grades prekindergarten through 8 for lead times of 1, 2, 5, and 10 years were 3.5, 4.9, 9.2, and 7.9 percent, respectively. For more information on mean absolute percentage errors, see table A-2 in appendix A.