Transcript 11.3 www.bea.gov

Implementation of SNA 2008 in
the 2013 Comprehensive Revision
of the U.S. National Income and
Product Accounts (NIPAs)
Marshall B. Reinsdorf
Workshop on SNA 2008 in Latin American National Accounts
Rio de Janeiro
September 17-18, 2013
www.bea.gov
Comprehensive revisions of NIPAs
▪ Comprehensive revision about every 5 years
▪ 14th comprehensive revision just happened in 2013
▪ Incorporate results of the 5-year economic census
and the benchmark input-output accounts
▪ Update the reference year for prices/quantities
▪ Opportunity to introduce major changes in
concepts, methods, and tables
▪ The entire time span back to 1929 is potentially
open for revisions
www.bea.gov
2
BEA’s implementation of SNA 2008
▪ The U.S. Bureau of Economic Analysis (BEA) had already
adopted some of SNA 2008 changes before 2013:
 Non-life insurance and reinsurance.
 Military fixed assets
Also had full sequence of accounts in Integrated
Macroeconomic Accounts.
▪ In 2013 comprehensive revision, BEA implemented the
major SNA changes affecting production and income:





Capitalization of research and development
Capitalization of costs of ownership transfer
Pension entitlements
Improvements to measure of FISIM
Entertainment, literary, and artistic originals (was in 1993 SNA)
www.bea.gov
3
For future implementation
▪ Several changes require development of new data
sources and methods and are on BEA’s work plan:
 Treatment of employee stock options
 Currently recorded by BEA when exercised
 SNA 2008 treatment is based on fair value and is
recorded between grant and vesting
 Recording of goods sent abroad for processing on a
change-of-ownership basis
 Merchanting to be reclassified as trade in goods
www.bea.gov
4
Borrowing/Investment from the
Integrated Macroeconomic Accounts
www.bea.gov
5
Holding Gains and Saving of Households
from the Integrated Macro Accounts
Saving and Holding Gains of Households as Percent of Household Disposable Income
80
70
60
50
40
30
20
0
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
10
-10
-20
www.bea.gov
net saving
holding gains/losses
6
0
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Holding Gains and Saving from the
Integrated Macro Accounts
Saving and Holding Gains of Households as Percent of Household Disposable Income
100
50
-50
-100
-150
net saving
www.bea.gov
holding gains/losses
7
Research and development
▪ Expenditures on R&D have the characteristics of fixed assets
and should be treated as investment:
 Ownership rights, long-lasting, used in production
▪ Previous treatment:
 Business R&D expenditures were classified as intermediate inputs
 R&D expenditures of nonprofit institutions and governments were
included in consumption expenditures
▪ New treatment:
 R&D expenditures by businesses, NPISH, and governments are
counted as fixed investment
 Depreciation of R&D added to consumption of fixed capital (CFC)
www.bea.gov
8
From R&D expenditures to GDP impacts
▪ Identify R&D investment
 Sum R&D input costs based on performer data
 Remove double-counting of software
 Include depreciation of other fixed assets used to produce R&D
 Assign investment to owning sector—usually funder of the R&D
 Business
 Nonprofit institutions serving households
 Government (federal and state and local)
 Deflate nominal investment
 Input-cost approach with a productivity adjustment
▪ Estimate R&D capital stocks by owner
 Create capital stocks with perpetual inventory method
 Model derives industry-specific depreciation rates from investment & profits
 For general government, based on useful service lives of technologies
www.bea.gov
9
Impact of capitalizing R&D in U.S.
▪ Total R&D investment for 2012: $417.7 billion
 2.6% of (revised) GDP
 Business R&D: $248.7 billion
 Formerly treated as intermediate spending
 Government & nonprofit R&D: $169.0 billion
 Reclassified from consumption to investment
▪ Impact on GDP revision for 2012: $396.7 billion
 2.5% of (previously published) GDP
 Business investment added to GDP: $248.7 billion
 Government & nonprofit—add CFC for R&D:
$148.0 billion
www.bea.gov
10
Top Private Business R&D-investing Industries
Percent of Private Business Investment in R&D
2007
1987
www.bea.gov
11
R&D: Quarterly estimates
▪ Private business R&D investment
 After 2007 - quarterly financial statements (Compustat)
 Before 2007 - wages and employment
▪ Federal R&D investment
 Interpolated based on trends in intermediate R&D services
▪ NPISH and state and local government R&D investment
 Interpolated as a smooth trend
www.bea.gov
12
Entertainment, literary, and artistic originals
▪ Original films, sound recordings, literary and
music compositions, and artwork that can be used
for the production and sale of copies
▪ Estimates for several types, including:





Motion pictures
Long-lasting television programs
Books
Music compositions and recordings
Miscellaneous artwork
▪ Treatment as fixed investment in way that is
similar to R&D, except entirely in private sector
www.bea.gov
13
Entertainment originals: Implementation
▪ Little data on production costs
 Except theatrical movies, pre-2007
▪ Value of Investment = Net Present Value of
Revenue (NPV) Minus Non-Artwork Cost
 Revenue data adjusted to include only revenue from new
works
 Net revenue is estimated by removing non-artwork costs
 Adjusted net revenue is multiplied by an “NPV factor” to
derive investment value of future revenue stream
 The discount rate is 7% real
www.bea.gov
14
Entertainment originals:
Implementation
▪ Prices of entertainment assets
 PPIs, CPIs
▪ Annual depreciation will follow a geometric
pattern, based on trends in NPV over time.





Theatrical movies: 9.3 %
Long-lived television: 16.8 %
Books: 12.1 %
Music: 26.7 %
Miscellaneous: 10.9 %
www.bea.gov
15
New NIPA Tables
Private Fixed Investment in Intellectual Property Products
www.bea.gov
NIPA series
Private fixed investment in intellectual property products
Software
Prepackaged
Custom
Own account
Research and development
Business
Manufacturing
Pharmaceutical and medicine manufacturing
Chemical manufacturing, excluding pharmaceutical and medicine
Semiconductor and other electronic component manufacturing
Other computer and electronic product manufacturing
Motor vehicles, bodies and trailers, and parts manufacturing
Aerospace products and parts manufacturing
Other manufacturing
Nonmanufacturing
Scientific research and development services
All other nonmanufacturing
Nonprofit institutions serving households
Universities and colleges
Other nonprofit institutions
Entertainment, literary, and artistic originals
Theatrical movies
Long-lived television programs
Books
Music
Other
Start date
1929
1959
1985
1985
1985
1929
1959
1959
1959
1959
1959
1959
1959
1959
1959
1959
1987
1987
1959
1959
1959
1929
1929
1949
1929
1929
1929
16
Real private intellectual
property products
[Percent change from preceding period, SAAR]
25
Private Investment in IPP
[Percent Change from Previous Quarter]
20
Percent
15
10
5
0
-5
-10
www.bea.gov
17
Costs of ownership transfer
▪ Old treatment
 Real estate brokers’ commissions on structures classified as fixed
investment
 Depreciated over the life of the structure (80 years)
▪ New treatment
 Commissions on structures and land, title fees, attorney fees, other
non mortgage related costs
 Depreciated over the typical holding period (12 years)
▪ Effects
 Increase GDP by the newly recognized investment (2007: $60 billion)
 Increase CFC more than investment (2007: $130 billion)
 Decrease net operating surplus (2007: $70 billion)
www.bea.gov
18
Costs of ownership transfer
www.bea.gov
19
Changes in Treatment of DB Pensions
▪ Accrual-based accounting replaces cash accounting
 Gives more accurate picture of compensation and sector saving.
 Actuarial methods, which depend on assumptions, must be used.
▪ New Pension Plan Sector
 Consistent with the Federal Reserve Board’s Flow of Funds Accounts.
 Part of Financial Corporations Sector
▪ New Tables
 National totals for defined benefit plans.
 Private, state & local government, and Federal DB plans.
www.bea.gov
20
Accrual-based measures of pensions
▪ New measures for defined-benefit (DB) plans
▪ Accrual-based accounting
 Matches income earned with related production
 Recognizes employer liabilities for promised pension benefits
 Replaces cash-based accounting
▪ Compensation of employees
 Deferred compensation
 Benefits accrued on services rendered in current period
▪ Interest
 Benefits accrued on services rendered in past periods
 Includes interest on unfunded actuarial liabilities
www.bea.gov
21
Defined benefit pension concepts
▪ Actuarial Liability (“Benefit Entitlement”)
 Actuarial value of accumulated benefit entitlements
 ∆AL = service cost + interest cost – benefits paid + effects of
assumption changes and plan amendments
 Service cost is also known as “normal cost”
▪ Unfunded Actuarial Liability (UAL) =
actuarial liability – plan assets
▪ Change in Plan Assets = contributions +
property income – benefits paid – admin. exp.
+/- holding gains/losses + net capital transfers
www.bea.gov
22
Challenges in designing the new table
▪ Guidelines of the 2008 SNA (table 17.8) and the
Flow of Funds Accounts have a pension plan sector
located in the financial corporations sector.
▪ Employer’s normal cost is compensation income.
▪ Don’t want to affect saving of financial corporations
sector.
 Saving by pension plans defined to equal zero.
 Dividend and interest income passed through to persons.
www.bea.gov
Challenges in designing the new table
▪ Gap between interest on the actuarial liability and property
income on plan assets is likely to occur because plans use
holding gains to fund benefits and have non-zero UAL .
▪ Assets that generate holding gains pay less income, so
property income < assumed interest rate  value of assets.
▪ If plan invests in assets that can be expected to generate
holding gains, shortfall in property income from assets vs.
the income implied by the assumed interest rate will be
termed “implied funding of benefits from holding gains” .
▪ Interest imputed on the loan from plan to the employer if
the unfunded actuarial liability (UAL) > 0, or on prepaid
contributions if UAL < 0.
www.bea.gov
New DB Pension Flows
Imputed Income from Defined Benefit Pensions
as percent of disposable personal income
2.0
1.5
1.0
0.5
0.0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
-0.5
-1.0
imputed employer contributions
www.bea.gov
imputed interest on unfunded pension entitlements
total imputed household income
25
DB pension flows in the NIPAs
Financial
assets
Benefit payments
and administrative
services
Monetary interest
and dividend income
Pension plan
Imputed interest cost of
plan’s funding gap (UAL)
Rerouted
employer
contribution
s
Employer
Labor services
Imputed interest
and dividend
income from assets
and employer
Contribution
supplements (equal
to imputed income
received) and direct
contributions
Employees and
former employees
Pension promises
(actual and imputed contributions as
compensation)
www.bea.gov
26
What do contribution lines in the table equal?
▪ Imputed employer contributions = admin expenses + service
cost–actual employer contributions–employee contributions
▪ Contributions includes household contribution supplements
and a negative imputation for administrative expenses, which
are recorded as implicit sales of services to households.
▪ Contributions = Gross accruals of benefit entitlements
excluding benefits funded by holding gains.
▪ Net change in benefit entitlements = Contributions – benefits.
www.bea.gov
Pension Table, Top Part
(research estimates for private plans in 2007)
Current receipts, accrual basis
Output
Contributions
Claims to benefits accrued through service to employers
Actual employer contributions
Imputed employer contributions
Actual household contributions
Less: Pension service charges
Household pension contribution supplements
Income receipts on assets (including plans' claims on employers)
Interest
Monetary interest
Imputed interest on plans' claims on employers (for the UAL)
Dividends
Current expenditures, accrual basis
Administrative expenses
Imputed income payments on assets to persons
Interest
Dividends
Benefit payments and withdrawals
www.bea.gov
Net change in benefit entitlements
223.3
9.8
147.5
81.4
67.1
23.3
0.8
9.8
66.1
66.1
33.9
37.6
-3.7
32.2
223.3
9.8
66.1
33.9
32.2
158.8
-11.3
Pension Table, Bottom Part
Cash flow
Actual employer and household contributions
Monetary income receipts on assets
Less: Benefit payments and withdrawals
Less: Administrative expenses
-30.9
67.9
69.7
158.7
9.8
Effect of participation in plans on personal income, saving, and wealth
Effect on personal income
Less: Effect on personal consumption expenditures
Equals: Effect on personal saving
Plus: Implied funding of benefits from holding gains on assets
Interest accrued on benefit entitlements
Less: Interest and dividend income received by plans
Equals: Change in personal wealth
Less: Benefit payments and withdrawals
Plus: Household actual contributions
Change in benefit entitlements including implied funding of benefits from holding gains
www.bea.gov
156.5
9.8
146.7
73.0
139.1
66.1
219.7
158.7
0.8
61.8
Effects of DB Plan Changes on Saving (research
estimates for 2007; $ billions)
Private
Business
State & Local
Governments
Federal
Government
Households
Saving,
DB pensions on
cash basis
270.7
12.2
-245.2
248.7
Revision in Saving
–19.6
–101.3
–35.1
+156.0
Revision, as percent
of disposable
personal income
–0.2
–1.0
–0.3
+1.5
Actual Revision
Published Data
–0.21
–0.95
–021
+1.3
www.bea.gov
Source data for private plans
▪ We add up variables on the ABO, normal cost, contributions,
benefits, and assets from almost 40,000 tax returns per year.
▪ Interest and dividend income of plans is estimated by
multiplying average rates of return by corresponding values
of assets.
▪ We use data sets from 2000 on (for early years extreme and
missing values were problems, and some plans were missing.)
▪ For pre-2000 years, we extrapolated back normal cost rate
using future benefits as a indicator.
▪ Reported numbers adjusted to reflect a common interest rate
assumption based on AAA corporate bond yields (5 percent
in recent years).
www.bea.gov
Data for state & local government plans
▪ For state & local government plans, we collected samples of
actuarial valuation reports covering 90% of assets and
membership back to 2000.
▪ We used membership data and estimates of normal cost
rates to extrapolate back to 1929 (beginning of time for the
NIPAs).
▪ Census Bureau will collect normal cost data in future, helped
by new reporting standards promulgated by GASB.
▪ Most of the reports use the Entry Age Normal method and
assume a high rate of interest; we adjusted them to ABO
method and to use same interest rate as we used for private
plans.
www.bea.gov
Data for federal government plans
▪ Actuarial reports go back to 1979 (civilians) or 1985 (military).
▪ Use PBO approach; legal funding targets are also based on PBO.
▪ Normal cost for each year, the PBO for 2013, and actual plan
expenses used as inputs into simulations. For older years, we
multiplied payroll by an estimated normal cost rate. Civilian
simulation incorporated plan rule changes in 1930, 1942, 1948, 1956
and 1969.
▪ We assumed that trust fund received contributions equal to
normal costs and earned interest on assets at the rate used by the
federal actuaries. It pays benefits and administrative expenses.
▪ Trust fund balance served as estimate of PBO, with upward
adjustments in 1970 for inflation surprise and in 2009-2010 for
interest rate decline that was not matched by inflation decline.
www.bea.gov
Accrued interest for unfunded
actuarial liabilities
220
200
180
160
Billions of dollars
140
120
100
80
60
40
20
0
-20
2002
2003
2004
2005
from private business
www.bea.gov
2006
2007
2008
from federal
2009
2010
2011
2012
from state and local
34
FISIM: Banking services
▪ Include only assets and liabilities with direct
customer contact
▪ Exclude expected credit losses from borrower
services
▪ Improve user cost estimate of depositor and
borrower services
www.bea.gov
35
Effect of smoothing and default
adjustment on borrower services
70000
60000
Published
method
$Millions (current)
50000
40000
30000
20000
Proposed
method
10000
0
www.bea.gov
Borrower services (no stabilization)
Borrower services (with reference rate stabilization)
Borrower services (stabilized, without default adjustment)
Borrower services, no adjustment
36