Job Protection Legislation and TFP Growth in OECD Countries'

1
Job Protection Legislation and TFP Growth in
OECD Countries.

• Discussion
• Federico Cingano

BI, November 2008
2
Question Does EPL slow down TFP growth?

• If EPL slows employment reallocation, it might
  prevent firms from fully exploiting existing
  growth opportunities.
• DID-like (Rajan-Zingales) approach ? Negative
  relationship
• ? i.e. high-EPL countries see slower TFP growth
  in industries with stronger need for labour
  reallocation (higher layoff rates)
• 73-pages, 23-tables paper, plenty of robustness
  checks and discussions.. Hard to think of
  something missing!

3
Empirical framework

• Year-by-year country-industry TFP growth
  (1982-2003)
• Interaction of country-level EPL with a measure
  of (frictionless) industry needs for
  layoffs/reallocation. Think of amount and
  relevance of regulation.
• Country and industry fixed-effects can be
  accounted for

4
Issue 1 measuring ?j

• Standard approach measuring industry
  requirements taking ?j from a benchmark, low
  regulated country (here, the US in 2000-2003..).
  Idea in those countries layoff/reallocation is
  unlikely to reflect frictions in the labour
  market
• But ?j might reflect idiosyncratic shocks (es.
  Fisman-Sarria, 2004)
• ? Measurement error attenuation bias?
• ? If shocks correlated with EPL can yield to
  estimate positive effects even when there are
  none (Ciccone and Papaioannou, 2006, 2008)

5
Issue 1 possible solution

• Estimate layoff/reallocation rates for a hp.
  frictionless economy using data on all
  countries ..
• .. and accounting for the fact that in countries
  with high employment protection, job reallocation
  might not just reflect industry characteristics

• Use estimates of aj and ?j to build alternative
  measure of frictionless industry-need for
  layoff/reallocation. The industry ranking might
  change significantly with respect to country
  benchmarking case. (fig)

6
Issue 2 methodology

• Standard approach When relying on
  benchmark-country weights, the benchmark country
  is normally excluded from the main regression (a
  problem here?)
• Standard approach provide, if possible,
  evidence of direct effect (i.e. of EPL on labor
  market flows)
• ? Note Red Tape costs shown to have strong
  effects of the speed of employment reallocation.
  Is EPL estimate robust to RT (?PMR)?

7
Issue 3 what is ? capturing?

• Do we estimate short- or long-run effects of EPL
  on TFP? (issue discussed in the paper)
• Interpretation problem arises because ?TFP is
  year-by-year and EPL does not vary much over time
  (? may capture lagged level effects)
• ? Why not try the standard cross-country
  cross-industry specification where growth rates
  are 20-year averages and EPL is
  beginning-of-period?

8
Example with industry job reallocation
Source Amadeus
9
Example
10
(No Transcript)