Notes

SMSA counties and account for approximately 73 percent of total
railroad employment. For the remaining counties, AAR's residual-
counties employment estimate in each State was disaggregated in
proportion to the railroad employment reported by counties in
the 1970 Census of Population. Estimates for the intervening
years were derived by averaging the biennial benchmark data.
The resulting employment series was used to allocate the State
totals of wages and salaries in the railroad industry. The 1975
AAR data were available for distributing the 1975 State totals.

Source: see note 22.

24. Here is the BEA description of their procedure:

Benchmark estimates of local government wages and salaries were
prepared for each county from the local government payroll data
reported in the 1967 and 1972 Censuses of Government. Estimates
for the intervening years were the products of straight-line
interpolation. The 1973-75 estimates were made in two parts.
For the 372 largest counties, the local government payrolls were
obtained from the annual Bureau of the Census publication, "Local
Government Employment in Selected Metropolitan Areas and Large
Counties." These 372 counties accounted for 71 percent of all
local government wages and salaries. The estimates for the
remaining small counties were made by extrapolating the 1972 bench-
mark estimates by population and using the extrapolated series
to distribute the residual State control totals (derived by sum-
ming the large county payrolls for each State and subtracting
the aggregates from the State control totals).

Source: see note 22.

25. The Bureau of the Census publishes its own payroll data by industry and county, in County Business Patterns. The data collection procedures are very well documented and eminently reasonable.

Why doesn't READ use County Business Patterns in preference to BEA? Three reasons are given:

County Business Patterns only covers about 80% of the economy.

• County Business Patterns does not provide estimates of non-labor income; BEA does. Wages and salaries from County Business Patterns turn out not to mix well with non-labor income from BEA.

• Until 1974, County Business Patterns employment counts were only for the two-week pay period including March 12.

On the first point, the part of the economy not covered by County Business Patterns is just the part of the economy where BEA does it all by allocation. The second point casts further doubt on the BEA non-labor income figures. The third point is a real limitation on the CBP data.

26. The procedure used by BEA is quite murky. Sample quotes from the documentation (note 22):

27.

The READ Model

That is, the State total for each income component, as taken
from the official State income series before adjustment for resi-
dence, is allocated to the counties of the State in accordance
with each county's proportionate share of the same or some
related series that is available on a county basis.

Almost 400 series of separate estimates go into the derivation
of the 30 line items shown in the published personal income
tables for the SMSA's and counties.

A telephone interview with a senior person at BEA produced the following kinds of responses:

We use the best procedures available.

Do you have any trouble with allocation?

You have to remember how bad the other data is.

Federal Energy Data System Analysis and Evaluation, IDEAMATICS, June 30, 1978.

31.

In READ, "replacement" investment is in fact allocated by payroll shares, but new investment is not (section 6).

A MODELER'S VIEW OF THE READ MODEL ASSESSMENT PROCESS

Frank Hopkins

Energy Information Administration

Department of Energy
Washington, D. C.

I. INTRODUCTION

This paper assesses the recent Regional Energy, Activity and Demographic (READ) model review conducted by a number of distinguished academicians and members of the Office of Analysis Oversight and Access, Energy Information Administration (EIA), Department of Energy (DOE). The paper, authored by the Director of the READ project, is intended to provide suggestions for improving the procedures for reviewing other models, which are expected to be commonplace in the future.

The discussion in this paper presumes that the reader is familiar with the general nature of the Mid-Range Energy Forecasting System (formerly PIES), the regional policy and forecasting problems involved with the interaction between energy and the economy, and the general structure of the READ model.

The remainder of the paper will be divided into seven sections that attempt to chronologically record the development of the review process. Section II will outline the intended role of the READ model in improving the analytical capability of EIA. Section III will discuss the purpose and cost and benefits of the review. The structure of the review will be outlined in Section IV. Section V will present an analysis of the major objections to the county-level READ model as preceived by the Review Committee. The READ model's staff response to modify the model by estimating it at an SMSA level will be discussed in Section VI. The responses of the Committee made at the last review meeting to the SMSA proposal, as preceived by the READ staff, will be summarized in Section VII. The conclusion will present the recommendations of the READ staff on improving future reviews.

II. EIA MID-RANGE MODELING SYSTEM INTERFACE

This section is designed to provide background information to enable the reader to appreciate the larger context of model development in EIA. The role of the READ model in improving EIA's applied analysis capability will be to incorporate feedback from the regional energy projections directly to regional economic projections. In addition, these regional economic projections help to determine regional energy demands through a set of Structural Econometric Energy Demand (SEED) models. Thus, for the first time, the EIA modeling system would embody an energy-economy interaction at the regional level.

This paper is a discussion of the assessment process
as preceived by the author. It is not intended to nor
does it represent a policy statement of the Department
of Energy