Labor Supply Topics

1
Labor Supply Topics

• David L. Dickinson
• Appalachian State University
• April 2006 GATE.

2
An experimental examination of labor supply and
work intensities. David L. Dickinson, 1999, JOLE

• Whats the question? (or paradox)
• Static labor supply model predicts positive
  Slutsky (compensated) substitution effects of a
  wage increase on work hours.
• Some empirical studies estimate negative
  compensated wage effects on work hours.
• How can we solve the puzzle?

3
Can theory be refined to explain the empirical
data better?

• Classical model posits that workers choose only
  work hours
• Implicit assumption is that work effort is fixed.
• A specification allowing hours and effort (while
  on-the-job) choices can generate negative
  substitution effects on hours of work if
  U23lt0the Combined Model.
• Another model makes effort predictions when hours
  of work is not a choicethe Intensity Model.
• Can experiments help select among theories?

4
Experimental System

• How can we generate data on leisure preferences?
• Induced valuation not helpful here.
• Some measurement of leisure choices necessary
• As such, we can never know exact underlying
  preferences, only whether data is consistent with
  hypothesized utility function parameters.
• Labor task cannot be enjoyable, since a test of
  the classical theory would not then be possible.

5
Labor system used

• Labor task is typing paragraphs for piece-rate
  wages.
• Why not hourly wages?
• Wage change decomposed into substitution and
  income effect over 3 experimental work days.
• Initial wage state must occur on first work day,
  but substitution/income effect days randomly
  order for different subjects
• Wage increases and wage decreases used.
• IOUs used since data on all days necessary.
• Inter- and intra-day learning a concern.
• Two institutions usedfixed hours and non-fixed.
• To test both Combined and Intensity Models

6
Results

• In Intensity Experiments, income-compensated
  substitution effects on work effort positive
  (theory supported)
• In Combined Experiments, aggregate substitution
  effects on hrs. worked positive, but about 30 of
  individual subjects?negative subst. effects
• Consistent with Combined Model (substitution of
  on- and off-the-job leisure)
• Results highlight empirical value of Combined and
  Intensity models, confirm that full labor supply
  response includes an accounting for work effort
  changes.
• Note results can discriminate between some
  classes of models, but the experiments still have
  limitations.