Preview of the 2018 Comprehensive Update of the Regional Economic Accounts

By David G. Lenze  |  Volume 98, Number 8  |  August 16, 2018

Download PDF

This fall, the Bureau of Economic Analysis (BEA) will release the initial results of the 2018 comprehensive, or benchmark, update of the Regional Economic Accounts. Updated statistics will be released for personal income, gross domestic product (GDP), and personal consumption expenditures (PCE). Comprehensive updates are usually conducted at 5-year intervals; the last comprehensive update of the regional accounts was released in September 2013.

Comprehensive updates and, to a lesser extent, annual updates provide the opportunity to introduce data revisions, changes in definitions, and updated statistical methods from the comprehensive update of the National Income and Product Accounts (NIPAs) and the Industry Economic Accounts;¹ to incorporate newly available and revised regional source data; and to introduce improvements in regional statistical methods.

The purpose of this update is to more accurately portray evolving state and local economies, to provide consistent comparisons over time and with the national and industry accounts, and to provide additional data or perspectives for users.

Initially, revised annual and quarterly state personal income statistics from the first quarter of 1998 to the first quarter of 2018 will be released in September; these statistics were compiled according to the North American Industry Classification System (NAICS). Revisions to earlier periods and to local areas will be released subsequently.

Some of the more important changes affecting personal income, GDP, and PCE are described in detail below.

Annual state personal income

Compensation of state and local government employees. The compensation of employees includes two components related to defined benefit pension plans: employer’s normal cost and pension service charges. Employers’ normal cost is normal cost net of actual household contributions.

As part of the comprehensive NIPA update, measures of normal cost for state and local government defined benefit pension plans will be based on the projected benefit obligation (PBO) method rather than on the accumulated benefit obligation (ABO) method. The PBO method is currently used for the federal government defined benefit pension plans, so the change will enhance the comparability of the two measures, Private sector defined benefit plans will continue to be presented on an ABO basis.

The state estimates of employers’ normal cost for 2000 to the present will be prepared in the same manner and with the same source data as the NIPA estimate.² Beginning with fiscal year 2014, state and local government pension plans began reporting normal cost (also known as service cost) in their comprehensive annual financial reports. The estimate is based on the entry age actuarial cost method, a type of PBO method. BEA will collect the actuarial data for a sample of pension plans whose membership represents approximately 85 percent of all pension plans and adjust the data to a common discount rate. (The state estimates for 1999 and earlier years are extrapolations using active membership in the pension plans as an indicator series).

Currently, the NIPA estimate of pension service charges is allocated to states in proportion to employers’ normal cost. Beginning with 2000, the administrative expenses data collected by the Census Bureau as part of the Survey of Public Pensions (the same source data as used for the NIPA estimate) will be used to prepare state estimates of pension service charges, which will then be reconciled with the NIPA estimate. The source data are by place of work, as is compensation data.

Personal interest income. Personal interest income includes an imputation for depositors’ services furnished without payment by financial intermediaries. Commercial banks, savings institutions, credit unions, and regulated investment companies often do not charge depositors an explicit fee for services such as check clearing and recordkeeping. These services are both a type of interest income in kind as well as a type of personal consumption expenditure (PCE). Currently, for the purposes of state personal income, the NIPA imputed value of depositors’ services received by persons is allocated to states using the BEA estimate of monetary interest as an indicator. As discussed below, the imputed value of depositors’ services will be prepared for state PCE using state-level deposits data. To enhance consistency with state PCE, these new estimates will be used for state personal income as well.

Workers’ compensation. The state estimates of state-administered workers’ compensation are based on the Census Bureau’s State Government Finances data. BEA adjusts the Census Bureau data from fiscal years to calendar years and from place of work to place of residence. The Census Bureau data include claims paid directly to, or on behalf of, injured beneficiaries for compensation of wages lost, medical care, rehabilitation, funeral expenses, and other eligible benefits. Currently, the residence adjustment is based on state-level wage outflow ratios. In recent years the residence adjustment appears to have introduced a bias towards high wage states such as New York. Beginning with the estimates for 2001, the residence adjustment of workers’ compensation will be based on state-level employment outflow ratios. These ratios will be calculated in the same fashion as the wage outflow ratios, but they will use employment from the Census Bureau’s commuting (journey to work) data rather than wages.

Reclassification of Nevada’s and New York’s payroll taxes. In 2004, Nevada began levying the Modified Business Tax, a tax on gross wages less employee health benefits of every Nevada business that is subject to the unemployment compensation law. In 2010, New York began levying the Metropolitan Commuter Transportation Mobility Tax, a tax on the payroll expense of employers and self-employed individuals engaging in business within the New York metropolitan commuter transportation district. The Census Bureau’s Annual State Government Tax Collections Survey includes these taxes in other state business taxes and state taxes, not elsewhere classified. Those items are the source for the NIPA estimate of state and local government other personal taxes, thus the taxes are currently being recorded as personal taxes. However, since payroll is the tax base for both taxes, beginning with 2007, the NIPAs will treat them as a tax on production and record them as other business taxes on production and imports. The same changes will be made to the treatment of these taxes in state personal income and state GDP, which use the same Census Bureau source data.

Reclassification of the net income of tax-exempt rural telephone cooperatives. Currently, BEA classifies the net income of tax-exempt rural telephone cooperatives (a component of nonfarm proprietors’ income) in the utilities industry. Beginning with 1998, the net income will be classified in the telecommunications industry.

Quarterly state personal income

State unemployment insurance compensation. Currently, quarterly state unemployment insurance compensation is estimated using monthly data from the Department of Labor’s Employment and Training Administration (ETA) on the amount of benefits paid to unemployed workers by state.³ Annual estimates are distributed to quarters using the seasonally adjusted ETA data as an indicator. The ETA data are also used as an indicator when extrapolating forward. This method introduces volatility into the quarterly estimates that reflect irregularities in the timing of reported benefits rather than changes in the number of beneficiaries or in the size of actual benefit payments. Beginning with the first quarter of 2008, the seasonally adjusted unemployment rate from the Bureau of Labor Statistics’ Local Area Unemployment Statistics (LAUS) program will be used as the indicator series. This change is consistent with the change recently made to the NIPA methodology.⁴

During periods of long and widespread unemployment, Congress often enacts temporary programs that expand benefit eligibility beyond what is allowed by the regular state unemployment insurance programs. Emergency Unemployment Compensation data, available by state from ETA, will be used to allocate a national special factor representing the unemployment compensation not otherwise accounted for by the methodology.

Wages and salaries. Currently, the NIPAs record the pay raises that federal government military and civilian employees often receive in January as a level increase in first quarter of the year. By convention, BEA includes these seasonal movements to better illustrate the effects of the federal employee pay raises. As part of the comprehensive update, the NIPA measures of federal wages and salaries will be revised to remove this seasonal effect from the first-quarter estimates. Since the NIPA measures control the quarterly state estimates of federal civilian and military wages and salaries, this seasonal effect will also be removed from them as well.

County personal income

Unemployment insurance compensation. Workers who are unemployed because of the adverse economic effects of international trade arrangements may be eligible to receive trade adjustment assistance. Currently, the state estimates of trade adjustment assistance, a component of unemployment insurance compensation, are allocated to counties using a 1986 distribution of benefits from the Department of Labor. Beginning with 2009, the state estimates will be allocated to counties using the number of workers covered by certified petitions. The annual county tabulations of these workers are provided by the Department of Labor.

Employee and self-employed contributions for government social insurance. Employees in several states contribute to the state unemployment insurance program and to the state temporary disability insurance program. These contributions are a component of employee and self-employed contributions for government social insurance. Currently, the state estimates of these contributions are distributed to counties using the adult civilian population as an indicator series. Beginning with 1969, the state estimates will be distributed to counties in proportion to the BEA estimates of wage and salary employment. Although employee contributions are a fixed percentage of wages, the taxable wage base on which the contributions are paid is generally low. Most employees are likely to contribute similar amounts and so employment is likely to be a better allocator than wages.

Major Source Data introductions for personal income

In addition to the usual set of source data that BEA incorporates in the annual updates of the regional economic accounts, BEA will introduce several other recently available sets of source data. The residence adjustment for state and county personal income will introduce new county wage outflow ratios based on a special tabulation of 2011–2015 journey to work (JTW) data from the American Community Survey. Currently, the wage outflow ratios are based on 2006–2010 JTW data, which were used to establish a benchmark for 2008. The 2011–2015 JTW data will be used to establish a benchmark for 2013. The ratios will be interpolated between the benchmark years, and the 2013 benchmark ratios will be held constant for subsequent years.

The American Community Survey 2011–2015 5-year estimates will also be incorporated in the county estimates of personal contributions for veteran’s life insurance and the county-level non-education component of state government payrolls of several states.

The state source data used to estimate the income of nonfarm tax-exempt cooperatives have recently become available and will be incorporated in the comprehensive update. For agricultural cooperatives, net income data by state for 2001 onwards have been obtained from USDA’s Rural Development Program. For rural electric cooperatives, patronage capital (or margins) data by state for power supply borrowers and for distribution borrowers for 2001 onwards have been obtained from USDA’s Rural Utilities Service.

Expanded industry detail. Currently, in the annual state GDP statistics, hospitals (NAICS 622) are combined with nursing and residential care facilities (NAICS 623). Beginning with 1997, the industry detail will be expanded to include those industries separately. The same data currently used to estimate the combined industries are available for each industry separately. This will enhance consistency with the industry detail provided by other BEA accounts, such as state personal income, which already break out hospitals as well as nursing and residential care facilities.

Improved estimates of rental of tenant-occupied and owner-occupied nonfarm housing. Currently, the state estimates of rental of tenant-occupied nonfarm housing are based on contract rent data from the 1-year public use microdata sample (PUMS) of the American Community Survey (ACS). Some of rental units include utilities in the contract rent. This is problematic because expenditures on utilities are estimated and reported separately in PCE.⁵ The presence of utilities in contract rents does not create a problem of double counting because the state estimates control to the NIPA estimate for tenant-occupied housing, which is based on different source data and is net of utilities. However, it distorts the state distribution of the rents paid. Beginning with the estimates for 2001, the PUMS contract rent data will be adjusted to remove an estimate of utilities before the data are used to estimate rental of tenant-occupied nonfarm housing by state. This adjustment, in addition, will improve the state estimates of imputed rental of owner-occupied housing, which are based on the tenant-occupied estimates.

Improvements to the measurement of financial services and insurance. Currently, the NIPA estimate of personal consumption of financial services and insurance is allocated to states using disposable personal income as an indicator series. Beginning with 1997, a larger set of source data and indicator series will be used to estimate the detailed components of financial services and insurance (table 1). The revised state methodology and data sources offer two major advantages: (1) the estimates will reflect more accurately cross‐state and cross‐time differences in households’ consumption of various financial and insurance services and (2) consistency between the state accounts and national accounts in the source data and methods used to estimate PCE will be enhanced.