Comments on Kerwin Charles and Melvin Stephens by HC120807091854


									Kerwin Charles and Melvin Stephens

   “The Level and Composition of
Consumption over the Business Cycle”

  Comments by Christopher Jencks
           June 9, 2005
             Main Findings
• Mean expenditures hardly changed
  between 1988 and 2000 for any income
  group in the Consumer Expenditure
  Survey (CEX).
• Expenditures in top three income quartiles
  were not affected by the business cycle.
• In the bottom income quartile, a 2 point
  swing in unemployment only changed
  expenditures by about 6 percent.
          Problem I: Trends
• Hard to believe that living standards were
  no higher in 2000 than in 1988.
• Comparisons between the CEX and the
  National Income and Product Accounts
  (NIPA) suggest that CEX underestimates
  spending in many categories.
• This problem may have gotten worse
  between 1988 and 2000, although CEX
  and NIPA definitions differ (medical care).
                 Quarterly expenditures in 2000 dollars: NIPA versus CEX
                    NIPA personal consumption expenditure per household

          $8,300       CEX total expenditure per consumer unit                   $8,200




          1988         1990        1992         1994         1996         1998    2000

           Problem II:
     Recessions have changed
• Historically, a recession meant both higher
  unemployment and lower consumption.
• But this pattern changed in the 1990s.
• The 1990-91 recession involved an
  unusually small drop in consumption.
• The 2001-02 recession involved no drop at
  all in consumption (although rise slowed).
            Mean Quarterly Personal Consumption Expenditure per
               Household: 1969 to 2004 in chained 2000 dollars









     1969      1974-75    1980-82            1991             2001 2004

          What’s going on?
In the 2001-02 GDP fell in only two quarters
and consumption never fell. Employment
leveled off because:
  More consumer goods were imported,
  Productivity rose
  Domestic investment fell.
Unemployment rose because:
  Net immigration remained high
  New native job seekers exceeded retirees
Why doesn’t higher unemployment still
       lower consumption?
Many trends predict that unemployment
should lower consumption more than in past
  • Savings rate is down.
  • Unemployment insurance coverage is
  • Unsecured borrowing fell from 1990 to
    1993 (Is this consistent with industry

Why doesn’t higher unemployment still
   lower consumption? -- (cont.)
 Asset values matter more than saving
    rates for consumption smoothing
 Secured borrowing may have risen in
    1991-92 and 2002-2003
 Is the two earner family protective? How?
 Could inter-family transfers have risen?
 Did macroeconomic policy change in
     ways that protected more consumers?
       Conclusions about paper
• Need more checks on validity of CEX trend data
  for expenditures. Check NIPA.
• Need data on whether CEX incomes fell in early
  1990s and rose late 1990s.
• Need to define “treatment.”
• Estimate effect of “treatment” on cash and
  noncash income, total expenditure, and types of
• Spell out how people protect themselves from
  downturns, and see why lowest quartile fares
  worst, if it does. Is it just low assets?

To top