The Impacts of Integration and Trade on Labor Markets: Methodological Challenges and Consensus Findings in the NAFTA Context

1
The Impacts of Integration and Trade on Labor
Markets Methodological Challenges and Consensus
Findings in the NAFTA Context

• Michael Abbott
• Labor Economist
• Commission for Labor Cooperation
• Washington, DC
• May 11, 2004

2
Presentation Agenda

• Methodologies used in this literature
• Non-methodological limitations and complexities
• Pre- and post-NAFTA research findings (trade,
  employment, wages and income inequality effects)
• Lessons for developing countries in the study of
  trade liberalizing effects

3
Methodologies

• Pre-NAFTA Studies
• (Forecasting)
• Linked Macro-economic Models (LMMs)
• Computable General Equilibrium Models (CGE models)

• Post-NAFTA Studies
• (Evaluation Methods)
• Partial Equilibrium Models
• Qualitative/ Quantitative Methods

4
Linked Macro-economic Models (LMMs)

• Forecasting of aggregate economic activity
• Rely upon historical data relationships from
  estimated equations and parameter estimates
• Effect difference between the baseline and
  revised (with policy change) forecast
• Two models are linked together
• Allows for simultaneous effect forecasts

5
Computable General Equilibrium (CGE) Models

• Allow for isolation of direct and indirect
  effects of a policy
• Simulate relationships (interactions) between all
  sectors of the economy specific sectors and
  industries can be analyzed
• Allow for assumptions regarding economic behavior
• Sensitivity analysis to gauge robustness of
  results on parameter estimates
• A lot of variation and specificity between
  individual CGE models

6
Partial Equilibrium Analysis

• Estimate effects of a policy on one or more
  variables by holding others constant
• High degree of flexibility of analysis
• Regression analysis is very prevalent
• Tend to focus on one or a small number of
  variables interactions not usually tested
• Allow for tests of a policy, rather than
  since a policy

7
Qualitative/Quantitative Methods

• Not confined to purely statistical relationships
  between variables
• Quantitative reasoning of secondary research is
  common
• Allows for a large breadth of analysis
• Can capture numerous effects of a policy
• Causal relations are not definitively proven
  effective for since NAFTA effects

8
Non-methodological Limitations and Challenges

• Evaluation difficult due to recency of policy
  (i.e. NAFTA)
• Data comparability problems
• Constructing a controlled experiment in a social
  science setting extraneous variables and
  surprise shocks
• Ascribing trade liberalizing success or failure
  to individual, or a small number of, indicators
• Research tends to focus on one country rather
  than all NAFTA-countries

9
NAFTA Effects on Trade

• Pre-NAFTA consensus net trade creating, with
  Mexico showing the most significant gains. Trade
  diversion a strong possibility
• Post-NAFTA results
• Increased exports from Mexico to the U.S.
• Small increase in exports from Canada to the U.S.
• Minimal effect on U.S. trade (exports and
  overall)
• Trade diversion evidence is mixed occurs in
  some sectors, but not in others

10
NAFTA Effects on Employment

• Pre-NAFTA consensus no consensus for the U.S.
  and Canada generally assumed that employment
  would increase at least moderately in Mexico
• Post-NAFTA results no consensus
• Results tended to depend upon the various inputs
  added and the assumptions factored into the
  models
• Disentangling of from since NAFTA employment
  effects is difficult

11
NAFTA Effects on Wages and Income Inequality

• Pre-NAFTA consensus no real consensus small
  effects on wages for the U.S.
• Post-NAFTA results most studies are
  qualitative/quantitative
• No real evidence of wage increases specifically
  due to NAFTA product prices used to infer
  effect of trade on wages
• Income inequality has increased, but has not been
  directly attributed to NAFTA in econometric
  studies
• Lack of three-country analysis on these topics

12
Lessons for Developing Countries Studying Trade
Liberalization Effects

• Larger and more trade open countries tend to be
  less affected by a single agreement
• There is no consensus in the economics community
  on a best fit methodology
• Determining the proper methodology depends upon
  the studys goal (i.e. forecasting or evaluation)
  and the availability and comparability of data

13
General Factors that Affect Results

• Data comparability
• Assumptions that are built into the original
  model
• Generalized causal statements regarding effects
  of NAFTA versus effects since NAFTA can be
  specious
• The appropriate elapsed time from a policy change
  to determining its effects is debatable

14
Checklist Prior to Conducting Research

• Ensure that data is comparable and accurate a
  reliable statistics agency is essential
• The presence of extraneous variables must be
  accepted
• External shocks will affect forecasts and
  results
• Know what it is you are asking for!