From the February 2000 SURVEY OF CURRENT
BUSINESSBEA's measure of personal income and the IRS measure of AGI are two widely used measures of household income. In general, personal income, which is prepared as an integral part of the national income and product accounts (NIPA's), is the more comprehensive measure. Personal income is a measure of the current incomes earned by households and by nonprofit institutions serving individuals, and thus, it is often used in assessing trends in consumer spending, saving, and investment. It includes income that is generally taxed, such as wages and salaries, income from rent, self-employment earnings, dividends, and interest; income that is partly taxed, such as social security benefit payments; and several types of income that are not taxed, such as tax-exempt interest and nontaxable transfer payments, including medicare, medicaid, and welfare benefit payments. Personal income includes income whether or not it is properly reported to IRS. In order to provide a comprehensive measure of personal saving, personal income also includes other types of income, such as employer contributions to private and government employee retirement plans, the investment income of these plans, and imputed income related to home ownership. Personal income excludes net gains from sale of assets, pension benefit payments, and personal contributions for social insurance.
AGI, on the other hand, is an income concept defined by tax law, so AGI consists only of taxable sources of income net of specific adjustments as reported on IRS Form 1040. Therefore, it excludes many of the types of income that are included in the BEA measure.
Although the two series are based on different concepts and serve different purposes, they are often used in conjunction with one another. In particular, personal income, which is available much sooner than AGI, is frequently used as an extrapolator for AGI (this article provides information that enables users to adjust the BEA measure to bring it closer in definition to AGI). In addition, the AGI gap is used as a rough indicator of the noncompliance by individuals with the Federal tax code.
The new and revised estimates presented in this article reflect the incorporation of the results of the comprehensive revision of the NIPA's that was released in October 1999, of updated estimates of AGI from the Statistics of Income Bulletin, and of improved estimates of the items used to partially reconcile the two measures./2/ As explained in the section "Sources of the Revision to the AGI Gap," most of the revisions to personal income reflected the incorporation of definitional changes that did not affect the AGI gap./3/
The preparation of BEA estimates of AGI, "BEA-derived AGI," begins with the NIPA estimates of personal income. Personal income consists of the current income received by persons from all sources—that is, from participation in current production and from both government and business transfer payments./4/ It is calculated as the sum of wage and salary disbursements, other labor income, proprietors' income with inventory valuation and capital consumption adjustments, rental income of persons with capital consumption adjustment, personal dividend income, personal interest income, and transfer payments to persons, less personal contributions for social insurance.
The IRS measure of AGI is defined as the sum of all the items of "total income" less a set of specific adjustments to total income that are authorized by legislation. Total income includes all income that is received in the form of money, property, and services and that is not expressly exempt from taxation; it excludes, for example, interest on tax-exempt State and local government bonds, voluntary contributions to thrift savings plans, and nontaxable social security benefit payments. The specific adjustments to total income include subtractions for contributions to individual retirement accounts and Keogh plans, for alimony paid, for moving expenses, and for several items related to self-employment income.
Tables 1 and 2 show the reconciliation between personal income and AGI, by type of income, for 1996–97. Personal income and AGI each include items that the other omits by definition. The reconciliation items that convert personal income to AGI are shown in two groups: First, those items that are included in personal income but not in AGI, and second, items included in AGI but not in personal income. A third group of reconciliation items, "intercomponent reallocations," reallocates certain income components to make the BEA and IRS estimates of AGI comparable by type of income. The specific items included in each of these groups are discussed in the appendix to this article.
The estimates of the BEA-derived AGI differ significantly from the IRS estimates of AGI. The "AGI gap" is the difference between the total BEA-derived AGI (line 22) and total IRS AGI (line 23), and the AGI gap for each type of income (line 29) is the difference between the BEA-derived AGI for that type of income (line 22) and the reallocated IRS AGI for that type of income (line 28). The percent distribution of the AGI gap by type of income is shown in line 30, and the relative AGI gap for each type of income, which is the AGI gap for that type of income (line 29) as a percentage of the BEA-derived AGI for that type of income (line 22), is shown in line 31.
The AGI gap results from several sources. First, there are errors in the source data used to estimate the personal income components./5/ Second, there are errors in the IRS measure of total AGI and its components because the estimates are based on a probability sample. Third, there are errors in reconciliation items because (1) reliable data are unavailable to estimate some known items, such as income earned by individuals who are not required to file income tax returns, (2) some of the source data used to estimate known items contain errors, and (3) some of the differences between the definition of personal income and AGI are unknown. Fourth, the estimates of BEA-derived AGI include both explicit and implicit adjustments for tax-return misreporting—that is, noncompliance. Explicit adjustments are made for the effects of tax-return misreporting on the source data used to prepare the estimates of wage and salary disbursements, nonfarm proprietors' income, royalty income, and personal interest income (line 32)./6/ Implicit adjustments are also embedded in the source data used for some components of personal income because the source data are from the payers of the income. The IRS estimates of AGI are based on unaudited tax returns that are not adjusted for misreporting. (However, the sample returns are edited for consistent statistical definitions and for incorrect or missing entries in order to make them consistent with other entries on the returns and with accompanying schedules.)
BEA believes that the explicit and implicit adjustments for misreporting account for a major part of the AGI gap. In 1997, the explicit adjustments accounted for $311.5 billion of the $630.3 billion gap. Thus, the AGI gap can be considered a rough indicator of noncompliance with the Federal tax code, and the relative AGI gap—the AGI gap as a percentage of the BEA-derived AGI—can be considered a rough indicator of the noncompliance rate in the reporting of income included in AGI./7/
Table 3 shows the estimates of the AGI gap for total income and for each type of income for 1959–97, and table 4 shows the relative AGI gap for total income and for each type of income for 1959–97. Over this period, the relative AGI gap for total income averaged about 11 percent: It declined from 10 percent in 1959 to a low of about 9 percent in the late 1960's, increased to a high of 13{\erm\char189} percent in 1984, and then moved irregularly downward to about 11 percent in 1997.
The relative AGI gap for wage and salary disbursements is the smallest among the types of income, primarily because income tax withholding at the source is required for wage and salary disbursements. The relative AGI gap for wage and salary disbursements averaged about 3 percent over the period; it declined from 3 percent in 1959 to a low of 1 percent in 1982 and then increased to about 5 percent in 1997.
The trends in the relative AGI gaps for nonwage incomes partly offset each other (see the addenda in table 4). For nonwage incomes subject to the requirements for filing information returns, the trend in the combined relative AGI gap is generally downward./8/ The combined relative AGI gap for nonwage incomes subject to the filing requirements declined from about 39 percent in 1959 to 14 percent in 1997./9/
The combined relative AGI gap for incomes not subject to the filing requirements (the second addenda item in table 4) averaged about 50 percent over the period; it increased from about 32 percent in 1959 to the highest level of about 80 percent in 1984, decreased to about 48 percent in 1990, and then increased to about 56 percent in 1997.
Table 5 shows the revisions to personal income, reconciliation items, BEA-derived AGI, and the AGI gap for 1959–96. The revisions to the AGI gap reflected the definitional and statistical changes made in the 1999 comprehensive NIPA revision.
In general, revisions to the AGI gap result from three sources: Revisions to personal income that carry through to the AGI gap, revisions to reconciliation items that are unrelated to the revisions to personal income or to AGI, and revisions to AGI that carry through to the AGI gap.
For 1996, the AGI gap was revised down $54.0 billion (line 29), reflecting upward revisions of $122.2 billion to personal income (line 1) and net revisions of $176.2 billion to reconciliation items (lines 2 and 10). The revisions to personal income that resulted from definitional changes did not affect the AGI gap, because these changes also resulted in offsetting revisions to reconciliation items./10/ The $54.0 billion revision to the AGI gap reflected downward revisions of $14.7 billion to personal income as a result of statistical changes that were not offset by revisions to reconciliation items and upward revisions of $39.2 billion to reconciliation items that were unrelated to the revisions to personal income. There was no revision to the IRS estimate of AGI (line 23).
Not all of the $122.2 billion revision to personal income carried through to the AGI gap because these revisions resulted in $136.9 billion of offsetting revisions to reconciliation items. Of the $122.2 billion, $114.2 billion was attributable to definitional changes that were entirely offset by revisions to reconciliation items./11/ The $14.7 billion difference between the revisions to personal income and the offsetting revisions to reconciliation items reflected downward statistical revisions to some components of personal income that were carried through to reduce the AGI gap./12/
The definitional change that resulted in the largest revision was the reclassification of government employee retirement plans, which added $108.2 billion to personal income in 1996 and which resulted in several offsetting revisions to reconciliation items./13/ As a result of the reclassification, employer contributions, which were previously excluded from personal income, are added to personal income (line 1) and to other labor income (line 4); dividends and interest received by these plans are added to personal income and to investment income of pension plans (line 6); personal contributions to these plans are no longer included in personal contributions for social insurance (line 11)—a component deducted in the calculation of personal income—and thus are no longer deducted in the calculation of personal income; and benefit payments by these plans are no longer included in government transfer payments, thus reducing personal income but increasing taxable pensions (line 13) by the amount of benefit payments.
The $39.2 billion upward revision to reconciliation items unrelated to the revisions to personal income included revisions to investment income of pension plans (part of line 6), investment income received by nonprofit institutions or retained by fiduciaries (line 7), part of other personal income exempt or excluded from AGI (line 9), and other types of income (line 15). The revisions unrelated to the revisions to personal income reflected the introduction of new and improved methodologies or the incorporation of newly available and revised source data for reconciliation items that are prepared separately from NIPA revisions only for reconciliation purposes.
Reconciliation items shown in lines 6, 7, and 9 are components of personal income, but they are derived in aggregate in the estimate of personal income. Similarly, reconciliation items shown in line 15 are components of AGI , but they are also derived in aggregate in the estimate of AGI. Separate estimates for these reconciliation items are prepared only for reconciliation purposes. Thus, revisions to the separately estimated reconciliation items resulting from the introduction of new and improved methodologies or the incorporation of newly available and revised source data are unrelated to the revisions to personal income or to AGI and so are carried through to the AGI gap
In general, the incorporation of the results of the 1999 comprehensive revision resulted in small downward revisions to the relative AGI gap for total income (line 31) for all years except for 1991–93. However, the trend in the relative AGI gap for 1959–97 was largely unchanged.
In this appendix, the reconciliation items shown in tables 1 and 2 are explained in detail. The first group of reconciliation items consists of the items that are included in personal income but not in AGI; the second group consists of the items that are included in AGI but not in personal income; and the third group consists of the intercomponent reallocation items that make the BEA and IRS estimates of AGI comparable by type of income.
The first group of reconciliation items (lines 3–9) consists of the portion of personal income that is not taxable and therefore is not included in AGI.
Most transfer payments to persons from governments and all transfer payments to persons from business (line 3) are nontaxable./14/ The taxable portion of transfer payments to persons from governments consists primarily of unemployment compensation benefit payments and a portion of social security and railroad retirement benefit payments./15/
Most of the NIPA category of "other labor income" (line 4) is nontaxable./16/ The nontaxable components of other labor income are employer contributions to pension and profit-sharing plans, private group health and life insurance plans, privately administered workers' compensation plans, and supplemental unemployment benefit plans. Employer contributions are nontaxable, but some benefit payments—such as pension benefit payments and supplemental unemployment benefit payments—are taxable; they are in lines 13 and 15, respectively. Taxable components of other labor income include judicial fees to jurors and witnesses, compensation of prison inmates, and marriage fees paid to justices of the peace./17/
Imputed income in personal income (line 5) is certain incomes imputed to persons according to NIPA definitions./18/ Personal income wage and salary disbursements include three categories of wages and salaries in kind, the largest of which is the value of food furnished to military personnel and to other employees. Farm proprietors' income includes the imputed net rental income of owner-occupants of farm dwellings; nonfarm proprietors' income includes the imputed income of persons who participate in the construction of their own housing (the margins on owner-built housing). Rental income includes the imputed net rental income of owner-occupants of nonfarm dwellings. Net interest includes imputed income for services furnished to persons without payment by financial intermediaries except life insurance carriers.
Life insurance carriers and pension plans are not defined as persons in the NIPA's, but their investment income is included in personal income in the year it is accrued (line 6). Investment income of life insurance carriers attributed to persons is included in personal interest income./19/ Investment income of pension plans—rents, dividends, and interest—is recorded as being received directly by persons in the corresponding components of personal income./20/
Both nonprofit institutions and fiduciaries are defined as persons in the NIPA's, and their investment incomes are included, but not separately identified, in nonfarm proprietors' income, rental income of persons, personal interest income, and personal dividend income (line 7). Only the portion of the income of fiduciaries that is distributed to beneficiaries is taxable to individuals; the portion retained by fiduciaries is taxable to the fiduciaries and is not reported on individual income tax returns. Thus, investment income of fiduciaries included in line 7 represents the portion of fiduciaries' investment income that is retained by the fiduciaries.
The differences between NIPA accounting and income tax accounting for incomes included in both measures (line 8) consist mainly of differences in the valuation of inventories and depreciation and of differences in the methods of accounting for interest on bonds and for the earnings of individual retirement accounts and Keogh plans.
The inventory valuation adjustment for nonfarm proprietors' income is the difference between the cost of inventory withdrawals valued at their acquisition (historical) cost and the cost of inventory withdrawals valued at their replacement cost (the concept underlying the NIPA's). This difference is an estimate of inventory profits, but with the sign reversed. These profits are excluded from personal income but are included in AGI.
The NIPA measure of depreciation (consumption of fixed capital) is based on consistent accounting and is valued at current replacement cost, whereas the IRS measure of depreciation is based on varying service lives and depreciation formulas and is valued at historical cost./21/ The adjustment in line 8 consists of the differences for proprietors' income and rental income of persons except for depreciation on owner-occupied dwellings. (The entire amount of the rental income on owner-occupied dwellings, which is a NIPA imputation, is included in line 5.)
Interest on U.S. savings bonds, State and local government bonds, and corporate bonds is included in personal income on an accrual basis but is mostly reported on a cash basis in AGI. Similarly, earnings of IRA's and Keogh plans, excluding capital gains, are included in personal income on an accrual basis but are included as benefit payments in AGI.
Other personal income exempt or excluded from AGI (line 9) consists mainly of voluntary contributions by employees to thrift savings plans, tax-exempt interest received by individuals, tax-exempt military pay and allowances, the small business corporation dividend adjustment, and statutory adjustments to AGI.
Employee contributions to thrift savings plans, primarily 401(k) plans, are included in personal income as wage and salary disbursements but are excluded from AGI.
Tax-exempt interest received by individuals, primarily on State and local government bonds, is included in personal income but is excluded from AGI.
Certain types of pay and cash allowances to members of the armed forces, such as allowances to defray a portion of the cost of subsistence or to assist in obtaining civilian housing as a substitute for government quarters, are included in personal income but are excluded from AGI.
The small business corporation (S corporation) dividend adjustment is the difference between S corporation distributions that are included in personal income as personal dividend income and S corporation distributions that are considered as dividends in AGI. In the NIPA's, these distributions to individuals are, in their entirety, treated as personal dividend income for all time periods. In AGI, they have been treated at different times as dividends, partnership income, or small business corporation income./22/
Statutory adjustments are specific adjustments to total income that are allowed as deductions in the calculation of AGI. For 1997, statutory adjustments consist of contributions to Keogh and self-employed simplified employee pension plans and to savings incentive match plans for employees, certain contributions to IRA's, one-half of self-employment tax, a portion of self-employed health insurance premiums, forfeited interest and penalties incurred by persons who prematurely withdrew funds from time savings accounts, alimony payments, medical savings account contributions, certain moving expenses, foreign housing exclusion, repayments of supplemental unemployment compensation, certain expenses of qualified performing artists, and the amount of jury duty pay reported on Form 1040 that was repaid to employers.
The second group of reconciliation items (lines 11–15) consists of the portion of AGI that is excluded from personal income.
Personal contributions for social insurance (line 11), which is subtracted in the calculation of personal income, is included in AGI.
Net gain from sale of property (line 12) includes net gain from sale of property held for personal use or investment (capital assets) and of property of a business nature (business assets).
Taxable pension benefit payments from pension plans (line 13), which are in AGI, are excluded from personal income because of the NIPA treatment of pension plans.
Small business corporation income (line 14) in AGI is the amount of distributions from small business corporations that is taxable to individuals as ordinary income. As mentioned before, the amount taxable to individuals as dividends is in the IRS measure of dividends. In the NIPA's, small business corporation income is part of corporate profits, and the distributions to individuals are treated, in their entirety, as personal dividend income.
Other types of income (line 15) consists of income of U.S. citizens living abroad for more than a year, supplemental unemployment benefit payments, and "other AGI items" excluded from personal income. Personal income excludes income of U.S. citizens living abroad for more than a year because they are considered nonresidents of the United States for NIPA purposes, whereas AGI includes their income./23/ U.S. citizens are generally taxed on their worldwide income regardless of the geographic sources of their income and regardless of how long they have been living abroad. AGI from abroad mainly consists of wages, but a small amount is nonfarm proprietors' income.
Supplemental unemployment benefit payments from company-financed funds are reported as wages in AGI.
Other AGI items excluded from personal income include withdrawals from individual retirement accounts and Keogh plans, alimony received, State income tax refunds, net operating loss, gambling earnings, prizes, awards, and sweepstakes winnings.
The third group of reconciliation items consists of "intercomponent reallocations." The types of income used for reconciliation purposes in tables 1 and 2 represent the types of income that are common for both personal income and AGI. However, certain components of personal income and of AGI do not fit into this income classification or are classified differently. These income components are reallocated, when possible, to make the BEA and IRS estimates of AGI comparable by type of income. The reallocations affecting personal income are shown in lines 17–21, and those affecting AGI are shown in lines 25–27. The reallocated personal income and AGI by type of income are shown in lines 22 and 28, respectively.
The fees components of other labor income (line 17) are reallocated from other labor income to wage and salary disbursements because they are largely reported as wages./24/
Partnership income retained by fiduciaries (line 18) is reallocated to farm and nonfarm proprietors' incomes. Partnership income retained by fiduciaries is part of "income retained by fiduciaries," which is prepared specially for this reconciliation (line 7). Partnership income requires a reallocation because this type of income is not a type of income used for reconciliation purposes.
Interest received by, but not related to business operations of, nonfinancial proprietors and partnerships (line 19) is reallocated from personal interest income to nonfarm proprietors' income for years prior to 1987. This interest, which was tabulated in AGI as part of the income of proprietors and partnerships prior to 1987, is treated in the NIPA's as personal interest income in all time periods.
Taxable interest received by individuals from regulated investment companies (mutual funds) is reallocated (line 20) from personal interest income to personal dividend income because IRS requires that it be reported as dividends. Only the portion that is taxable to individuals is reallocated; interest paid by mutual funds to pension plans, IRA and Keogh plans, and holders of State and local government securities is tax-exempt. Because of difficulties encountered in allocating total interest paid by regulated investment companies to these different categories, this estimate of the amount reallocated from interest to dividends must be regarded as an approximation; the separate estimates of BEA-derived AGI for interest and dividends are therefore less reliable than those for the combination of the two./25/
Taxable disability income payments received prior to reaching minimum retirement age (line 21) are reallocated from taxable pension benefit payments to wage and salary disbursements because these payments are reported as wages on tax returns but are treated as pension benefit payments in personal income.
The AGI components that require a reallocation are estate or trust income (line 25), partnership income (line 26), and other reallocations (line 27). Estate or trust income is not a type of income that is used for reconciliation purposes; this income is reallocated to farm proprietors' income, nonfarm proprietors' income, and rental income of persons. Partnership income is reallocated to farm and nonfarm proprietors' incomes because this income is not a type of income that is used for reconciliation purposes.
Other reallocations (line 27) includes several reallocations that are not shown separately and are not applicable in recent years. For certain AGI components, the AGI classification has changed over the years, and the earlier classification must be adjusted to conform with the present AGI classification and with types of income used for reconciliation purposes. IRS business or profession net profit prior to 1963 is reallocated between farm and nonfarm proprietors' income because IRS did not tabulate farm proprietors' income separately. For 1958–65, the dividends distributed by small business corporations are reallocated from partnership income to personal dividend income; these dividends were tabulated by IRS as partnership income because the two types of business organizations were treated alike under tax laws. For 1957–65, Form 1040A wages that were not subject to withholding were tabulated by IRS as other income; this amount is reallocated to wages. For 1964–65, estate or trust income was tabulated as part of other income in AGI; this amount is reallocated to farm proprietors' income, nonfarm proprietors' income, and rental income of persons. For 1961, wage earners who had $200 or less of dividends and interest could report the combined amount as a single figure; this amount is reallocated to personal interest income and personal dividend income.
Footnotes:
1. Annual and quarterly estimates of BEA personal income are published monthly in table 2.1 of the national income and product accounts (NIPA's) in the section "BEA Current and Historical Data" of the SURVEY OF CURRENT BUSINESS (monthly estimates are shown in table B.1). Estimates of IRS AGI are published annually in Statistics of Income—Individual Income Tax Returns. The estimates of the relationship between total personal income and total AGI are presented annually in NIPA table 8.28, most recently for 1995–97 in this issue. All the estimates are available on the BEA Web site at <www.bea.doc.gov> and on the STAT-USA Web site at <www.stat-usa.gov>. The reconciliation by type of income for 1959–97 is available on request; for information, write to the Government Division (BE–57), Bureau of Economic Analysis, U.S. Department of Commerce, Washington, DC 20230. Estimates for 1947–58 will be available in April of this year.
2. For additional details of the 1999 comprehensive revision, see Eugene P. Seskin, "Improved Estimates of the National Income and Product Accounts for 1959–98: Results of the Comprehensive Revision" SURVEY 79 (December 1999): 15–43. For AGI data, see Internal Revenue Service, Statistics of Income Bulletin (Washington, DC: U.S. Government Printing Office, Spring 1999).
3. For additional details about the effects of the definitional changes on personal income and its disposition, see table 15 in Seskin, "Improved Estimates," 17.
4. "Persons" in the NIPA's consists of individuals, nonprofit institutions that primarily serve individuals, private and government employee retirement plans, and private trust funds.
5. AGI data are used only for the estimates of nonfarm proprietors' income and royalty payments in rental income of persons. For these components of personal income, the AGI gap does not result from errors in the source data, because the same errors are in the personal income components. The principal source data used to prepare other components of personal income are not based on AGI data. For additional detail, see "Updated Summary NIPA Methodologies,"SURVEY 78 (September 1998): 14–35 and the text on the CD-ROM State Personal Income, 1929–97; this information is also available on BEA's Web site. For changes in methodologies introduced in the 1999 comprehensive revision, see Seskin, "Improved Estimates,"37–39.
6. The major source data for these adjustments are the 1988 Taxpayer Compliance Measurement Program and Census Bureau "exact-match" files for 1990. For additional information about the calculation of these adjustments, see Robert P. Parker, "Improved Adjustments for Misreporting of Tax Return Information Used To Estimate the National Income and Product Accounts, 1977," SURVEY 64 (June 1984): 17–25; "The Comprehensive Revision of the U.S. National Income and Product Accounts: A Review of Revisions and Major Statistical Changes," SURVEY 71 (December 1991): 39–40; and "Improved Estimates of the National Income and Product Accounts for 1959–95: Results of the Comprehensive Revision," SURVEY 76 (January/February 1996): 24–25.
7. The income items that are excluded from personal income—such as net gains from the sale of assets, income from small business corporations, and alimony—are not adjusted for misreporting; thus, misreporting of these items may also contribute to the AGI gap.
8. Personal dividend income and personal interest income are the primary examples of nonwage incomes subject to the requirements for filing information returns. A combined AGI gap for personal dividend and personal interest income is also shown in tables 3 and 4 because of the difficulty in recent years of accurately estimating separate gaps for these incomes. The estimation difficulty is largely related to the types of income received by individuals from mutual funds and from private noninsured pension plans. The taxable portion of interest received by individuals from mutual funds is reallocated from personal interest income to personal dividend income (line 19); thus, the estimation difficulty affects both income gaps. In addition, the allocation of income received from private noninsured pension plans (part of line 6) between interest and dividends is based on incomplete data. Another estimation difficulty relates to possible misreporting by individuals of income from mutual funds on their tax returns. Although the IRS instructs individuals to report the income as dividends, some may have inadvertently reported it as interest.
9. Beginning with 1984, taxes have been withheld on taxable pension benefit payments unless the recipient elects not to have the tax withheld and on interest and dividends if the recipient fails to furnish a correct taxpayer identification number or has interest or dividends that were underreported on past returns.
10. Although the definitional changes per se did not result in revisions to the AGI gap, statistical changes unrelated to the definitional changes were carried through to the AGI gap. Statistical changes update the estimates to reflect the introduction of new and improved methodologies and the incorporation of newly available and revised source data.
11. For additional details, see table 15> in Seskin, "Improved Estimates," 27.
12. For example, rental income of persons for 1996 was revised down $21.7 billion with $3.9 billion offsetting revisions to reconciliation items. The difference ($16.6 billion) was statistical revisions that were carried through to reduce the AGI gap. Nonoffsetting statistical revisions were also in other components of personal income.
13. Government employee retirement plans, which were previously classified as social insurance funds within the government sector, are treated similarly to private pension plans. For more detail, see Seskin, "Improved Estimates," 26–29. Other definitional changes that resulted in revisions to personal income and to reconciliation items are the modification of the treatment of private noninsured pension plans; the redefinition of dividend payments by regulated investment companies (RIC's) to exclude distributions that reflect capital gains income; the redefinition of the value of imputed services of RIC's; and the reclassification of directors' fees. As mentioned in the text, these definitional changes did not affect the AGI gap, because the revisions to personal income from these changes were matched by offsetting revisions to reconciliation items.
The following reconciliation lines include revisions resulting from definitional changes: Nontaxable transfer payments (line 3), other labor income except fees (line 4), imputed income in personal income (line 5), investment income of pension plans (part of line 6), personal contributions for social insurance (line 11), gains, net of losses, from sale of property (line 12), and taxable pensions (line 13).
14. NIPA table 3.12 shows government transfer payments to persons. Government transfer payments to persons include benefit payments from social insurance funds—such as old-age, survivors, and disability insurance (social security), hospital insurance, supplementary medical insurance, and unemployment insurance—and from certain other programs. NIPA table 8.16 shows business transfer payments to persons. Business transfer payments to persons consists primarily of liability payments for personal injury and of corporate gifts to nonprofit institutions.
15. In the 1999 comprehensive revision, the treatment of government employee retirement plans was changed to make it more similar to the treatment of private pension plans. This definitional change resulted in a shift of the savings associated with government employee retirement plans from the government sector to the personal sector. Previously, government employee retirement plans were treated as social insurance funds within the government sector, and benefits paid by the plans were treated as government transfer payments to persons. For additional details, see Brent R. Moulton, Robert P. Parker, and Eugene Seskin, "A Preview of the 1999 Comprehensive Revision of the National Income and Product Accounts: Definitional and Classificational Changes," SURVEY OF CURRENT BUSINESS 79 (August 1999): 11–12.
16. NIPA table 6.11C shows other labor income by industry group and by type.
17. In the 1999 comprehensive revision, corporate directors' fees paid to "outside directors," which were previously in other labor income, were reclassified to nonfarm proprietors' income. Outside directors are directors who are not employees of the company on whose board they serve. Directors' fees paid to employees who serve on their company's board of directors are classified as wages and salaries.
18. Employer-paid health and life insurance premiums, which are treated as an imputation in the NIPA's, are excluded from line 5; these premiums are included in line 4. NIPA table 8.21 (line 53) shows the imputations that are included in personal income.
19. NIPA table 8.20 (line 53) shows imputed interest attributed to persons from life insurance carriers.
20. For additional details, see Moulton, Parker, and Seskin, "Definitional and Classificational Changes," 12–13.
21. The NIPA methodology for depreciation reflects empirical evidence on prices of used equipment and structures in resale markets, which has shown that depreciation for most types of assets approximates a geometric pattern. See Robert P. Parker and Jack E. Triplett, "Preview of the Comprehensive Revision of the National Income and Product Accounts: Recognition of Government Investment and Incorporation of a New Methodology for Calculating Depreciation," SURVEY 75 (September 1995): 39–41, and Barbara M. Fraumeni, "The Measurement of Depreciation in the U.S. National Income and Product Accounts," SURVEY 77 (July 1997): 7–23.
22. For many of the years prior to 1983, IRS instructed taxpayers to report actual distributions to individuals as dividends on Schedule B (Form 1040) and to report the retained earnings as part of supplemental income on Schedule E (Form 1040), which was tabulated as "small business corporation income" in AGI. Beginning in 1983, individuals were instructed to report only the distributions from pre–1983 accumulated earnings and profits as dividends on Schedule B (Form 1040) and to report other distributions as supplemental income on Schedule E (Form 1040).
23. In the NIPA's, U.S. citizens who reside outside the United States for less than 1 year are considered U.S. residents, but those who reside outside the United States for 1 year or more are, with certain exceptions such as diplomats and members of the armed forces, considered nonresidents. Similarly, aliens who reside in the United States for less than 1 year are considered nonresidents in the NIPA's. Personal income excludes income of NIPA nonresidents, but a reconciliation is not needed because AGI also excludes their income. AGI is based on a sample of individual income tax returns that is drawn from filers of forms 1040, 1040A, 1040EZ, and 1040PC, including electronic returns; the sample excludes filers of forms 1040NR and 1042, which are used by NIPA nonresident aliens to report their income.
24. See the previous discussion of other labor income (line 4) in the "BEA Estimate of AGI" section.
25. Several statistical problems cause the separate estimates of the AGI gap for interest and dividends to be less reliable than the estimates for the combination of the two. See footnote 8.