Wacziarg: Measuring the Dynamic Gains from Trade

1
Wacziarg Measuring the Dynamic Gains from Trade

• By Alan Fiedorek and Jenya Parkman

2
Introduction

• Past literature on trade has found a positive
  relationship between openness and growth
• Rodrik and Rodriguez (2000) disputed the positive
  association because of problems measuring
  openness

3
Goals of this paper

• To improve the existing measures of trade policy
  openness
• To define the channels through which trade
  openness affects growth

4
Why trade can be good

• Technological spillovers and international
  transmission of knowledge
• Allocative efficiency (specialization according
  to comparative advantages)
• Bigger market ? economies of scale
• Leads to less distortionary domestic policies and
  disciplined macroeconomic management

5
Why trade can be bad

• Comparative advantage leads some countries to
  specialize in dead-end industries
• More trade necessitates bigger government and
  higher taxesThe explanation appears to be that
  government consumption plays a risk-reducing role
  in economies exposed to a significant amount of
  external risk. Rodrik (1998)

6
Wacziargs Innovative Ideas

• New measure of trade openness Based on a
  weighted average of tariff revenues, nontariff
  barriers and an indicator of overall outward
  orientation
• Explicit links between trade and growth
• Equations for six specific channels through
  which trade effects growth

7
Government Policy Channels

• Openness leads to macroeconomic policies that
  encourage stability, so domestic firms can
  compete on a global market.
• Openness may have ambiguous effects on the size
  of government

8
Allocation and Distribution Channels

• There are lower degrees of price distortion in
  open economies. Price distortion has adverse
  effects on factor accumulation and growth.
  Easterly (1993) Distorted input prices cause
  inefficiency in production while higher black
  market exchange rates discourage investment
• Trade openness encourages factor accumulation
  through creation of a bigger market and
  importation of previously unavailable capital
  goods and recent technologies

9
Caveat

• The author asserts that theres a direct
  tautological link between increased investment
  and growth. However, Robert Solows growth theory
  emphasizes that growth in technology (Á/A g) is
  what drives economic growth.

10
Technological Transmissions Channel

• If open economies are more exposed to a worldwide
  stock of productivity-enhancing knowledge, then
  trade openness can affect growth and convergence
  through technology transmissions
• Trade openness has an ambiguous effect on Foreign
  direct investment (FDI). Study by Harrison and
  Revenga (1995) suggests that open economies
  attract more FDI than closed ones. FDI increases
  growth through imports of capital goods and
  diffusion of knowledge and expertise

11
Empirical Methodology

• Three-stage least squares regression
• The model consists of an equation for the growth
  of per capita income, one for determining the
  nature of trade policy, and six channel equations
  describing the effects of trade policy on several
  growth determining variables (398)

12
The Equation

• ?/y Á/A a(?/k) ß(?/h)
• Growth is driven by growth in technology, and
  per-capita growth in physical and human capital
• The 6 channel equations attempt to isolate the
  effect of trade policy on these determinants of
  growth

13
The Data

• 57 countries 21 OECD, 13 Asia (China excluded),
  12 Latin America, 11 Africa
• Four five-year periods from 1970-1989
• Endogeneity bias is a concern? author instruments
  for every endogenous variable appearing as a
  regressor

14
Measurement Issues

• 3 categories of existing measures of trade
  openness Outcome measures, Policy indicators,
  Deviation measures.
• Outcome measures arent good because they suffer
  from endogeneity bias and dont reflect policy
  attitudes. Since outcome measures have been
  discredited, we must choose between the other
  two.

15
Deviation Measures

• Measures difference between observed trade volume
  and predicted free-trade volume to determine how
  restrictive the trade regime is
• 3 problems predictions are likely to be
  incomplete and inaccurate, determinants may be
  correlated with policy, and white noise
  disturbance term.

16
Policy Indicators

• Institutional features reflect a countrys
  position on trade.
• Problems are endogeneity problems with growth,
  limited data availability, dont reflect possible
  black-market loopholes, and weakly correlated
  among themselves
• However, this can be overcome by combining the
  variation in several measures to obtain an
  indicator of trade openness

17
The Trade Policy Openness Index

• Wacziargs index is based on the impact that a
  countrys trade policy openness has on its trade
  (importsexports) as a ratio of GDP
• Trade to GDP ratio can be seen as a result of
  policy, factor endowment and gravity determinant
  variables. The author controlled for the effect
  of the factor endowment and gravity determinant
  variables to isolate the variation in trade among
  countries attributable to a variety of trade
  policy measures.

18
Components of the Index

• Tariff barriers Uses the share of import duty
  revenues in total imports IMF
• Nontariff barriers Unweighted coverage ratio for
  pre-Uruguay Round time period published by
  UNCTAD. Weight in overall index reduced because
  of difficulties measuring these barriers.
• Liberalization Status Sachs and Warner variable
  used to account for time variations in nontariff
  barriers that are unaccounted for due to data
  unavailability.

19
Sachs and Warner variable

• Countries classified as closed if they failed one
  of these tests
• 1. it had average tariff rates higher than 40
  (TAR)
• 2. its nontariff barriers covered on average more
  than 40 of imports (NTB)
• 3. it had a socialist economic system (SOC)
• 4. it had a state monopoly of major exports
  (MON)
• 5. its black market premium exceeded 20 during
  either the decade of the 1970s or the decade of
  the 1980s (BMP).

20
Criticisms of Sachs and Warner variable

• Rodrik and Rodriguez (2000), argue that this
  variable is constructed to be conducive to
  finding a positive effect of openness on growth
  because much of the variation is attributable to
  the black market premium on the exchange rate and
  the state monopoly of exports
• The Sachs-Warner measure is so correlated with
  plausible groupings of alternative explanatory
  variablesmacroeconomic instability, poor
  institutions, location in Africa--that it is
  risky to draw strong inferences about the effect
  of openness on growth based on its coefficient in
  a growth regression. (RR 2000)
• Therefore, Wacziarg presents his data with and
  without this liberalization status variable.

21
Trade Shares Regressions

• Trade Policy 1 -34.73Import Duty Share -
  0.22Nontariff Barriers 11.26Liberalization
  Status
• Trade Policy 2 -60.91Import Duty Share
  0.24Nontariff Barriers

22
Summary Statistics for Growth and Openness

• These correlations suggest that the relationship
  between trade policy openness and growth, if any,
  will be conditional on other determinants (408)

23
Measurement of Channel Variables 1

• Uncontroversial and precise
• 1. FDI
• 2. Government consumption
• 3. domestic investment rate
• all measured as a share of GDP

24
Measurement of Channel Variables 2

• Quality of macroeconomic policy is measured by an
  average of the level of public debt and
  government deficit both as percentages of GDP,
  and growth of M2 net of total real output growth
• Technology transmission is approximated by the
  share of manufactured exports in total
  merchandise exports (imperfect proxy alert)
• Price distortions are measured by the black
  market premium on the official exchange rate

25
Correlations (as expected)

• Trade policy is positively related to FDI, macro
  policy quality, manufactured exports, and
  domestic investment ratio
• Each of these is positively correlated with
  growth
• Trade policy is negatively related to the black
  market premium, and government size
• Each of these are negatively associated with
  growth

26
Empirical Results
3.3 -9.2 3.3 64 20 18 100
27
Result from Trade Policy 1

• An 8.5 ( 1 st. dev.) increase in trade policy
  measure results in a 0.601 increase in annual
  growth.

28
Empirical Results
-10 -30.5 -0.6 86.1
74.6 -19.3 100
29
Result from Trade Policy 2

• An 8 ( 1 st. dev.) increase in trade policy
  measure results in a 0.246 increase in annual
  growth.
• Relationship between trade policy and growth is
  cut by more than half.

30
Robustness to the Specification
31
Robustness to Time Coverage

• To account for the possibility that the results
  are only significant due to one particular 5-year
  period, the author runs 4 regressions excluding
  each of the periods
• Includes data for 1990-1992 to widen the time
  span
• Uses 10-year periods to reduce the impact of
  short-term variability

32
Other Possible Channels
Human Capital
Income Inequality
33
Unconditional effect and residual test

• Direct correlation between trade openness and
  growth (omitting all channels) is highly
  significant and of the same magnitude.
• The residual effect of trade policy is positive,
  but not significantly different from zero
  instilling confidence in the exhaustiveness of
  the trade channels

34
Conclusion

• Trade openness affects growth mainly by raising
  the ratio of domestic investment to GDP
• FDI and quality of macroeconomic policy also have
  significant effects
• Future research should strive to improve measures
  of technology transmission and price distortions

35
Problems

• Manufactured exports are a terrible proxy for
  technology transmission
• Lots of collinearities between the different
  channels make it unclear whether the results are
  accurately reflecting the effects of these
  channels
• Only 2 of 6 channels are statistically
  significant when Sachs and Warner variable is
  omitted

36
Questions

• What would be a better proxy for technological
  transmissions?
• Through what other channels do you think trade
  affects growth?
• Which channels do you think are particularly
  interconnected?
• What are the policy implications of this paper?
• Do you think that data from the last twenty years
  would change the findings of this paper?