Transcript Is the Economy Doing as Well as the Numbers Suggest?

Is the Economy Doing as Well as the
Numbers Suggest?
Sage Advisory Services Fall 2007 Conference
Raymond W. Stone
Stone & McCarthy Research Associates
September 27, 2007
Nature of Economic Data
Economic Data Are Largely Derived from Surveys
In Contrast to:
Data points from Physical Sciences Derived from
“Controlled Experiments”
Misplaced Emphasis?
• Financial Press/Talking Heads attach too much
emphasis to the latest release
• Economists all too often treat economic data, as
“hard numbers”
• Policy-Makers can be misguided by deficiencies
in the data
Consider the Establishment or
Payroll Survey of Employment
• Includes: monthly inference drawn from a sample
of about 330,000 establishments accounting for
about 30% of all employment
• Includes: adjustment for net payroll changes
outside of sample due to births and deaths of
businesses
• Adjusted for recurring seasonality, based on the
history of recent years
Benchmark Revisions
• Once a year the BLS realigns monthly estimated
payroll growth to Unemployment Insurance
records
• On October 5 the BLS will release a preliminary
estimate of the March 2007 benchmark revision
• The problems with the payroll data go beyond the
magnitude or direction of the benchmark revision
Recall Record +752,000 March
2006 Benchmark Revision
Ma rch 2 0 0 6 Be nchm a rk Re v ision By I ndust ry
( 0 0 0 s)
Government
Other services
Leisure and hospitality
Educ & health services
Prof &Business Services
Financial activities
Information
Trade,Transportation & Utilities
M anufacturing
Construction
250
225
200
175
150
125
100
75
50
25
0
- 25
M ining &Natural resources
March 2006 Revision Skewed
Towards a Few States
Observations from 2006
Revisions
• Largest revisions were in states with a high incidence of
illegal or unauthorized workers
• Largest revisions were in industries wherein a high
incidence of unauthorized workers are employed
• From a separate BLS survey, about 45% of all
employment growth during the period was among foreignborn workers, mostly Hispanic
• Wave of legislation aimed at restricting illegal workers and
increasing penalties on employers that knowingly hire
illegal workers
Inferences From 2006 Revisions
• Payrolls understated, but payroll growth
overstated
• Much of the job growth during the period reflected
illegal workers going from “off-the-books” to “onthe-books”
• Wage & Salary Growth overstated
• Productivity understated
• Unit labor Costs overstated
Payrolls Understated, Payroll
Growth Overstated
Disconnects in Current Payroll Data…How
About those Residential Construction Jobs?
Food Service Payrolls Out of Sync with
Expenditures at Food Service Outlets
Food Service Payrolls as Function of
Spending
9800
Above Tr end- - >
9600
9400
9200
9000
8800
8600
8400
8200
340000
360000
380000
400000
420000
440000
March 2007 Benchmark Revision should be
downward
(to be released on October 5)
But even if the Benchmark Revision is not
negative, Payroll Growth is still overstated
• The shifting of illegal workers from “off-thebooks” to “on-the-books” renders an increase in
the rolls of unemployment insurance eligibility,
but really isn’t an increase in employment
• Overstatement of Payroll Growth renders an
overstatement of Personal Income Growth
• Overstatement of Payroll Growth renders an
understatement of productivity growth
Policy Decisions Misled by Payroll and
Income Data?
(August 7 FOMC minutes)
In their discussion of monetary policy for the
intermeeting period, Committee members again
agreed that maintaining the existing stance of
policy at this meeting was likely to be consistent
with the overall economy expanding at a moderate
pace over coming quarters and inflation pressures
moderating over time. The expansion would be
supported by solid job gains and rising real
incomes that would bolster consumption
What About Construction
Productivity?
X-Construction, Productivity Stable
Potential GDP, the economy’s “speed-limit”
may be a little faster than the Fed thinks