Skip to main content

Part of the book series: Studies in Empirical Economics ((STUDEMP))

  • 192 Accesses

Abstract

In Section 2.2, we derived the value version of the HOV model, which allows for different factor productivities and factor returns across countries, while the technology parameters \(\theta _i^h\), the cost shares of factor h in industry i, are taken to be identical across countries. The HOV equation for a particular country j is (according to (2.18)):

$${\Theta ^T}{t^{jv}} ={W^j}{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\leftharpoonup}$}}\to {v} ^j}\sum\limits_j {{W^j}{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\leftharpoonup}$}}\to {v} }^j}} $$
((1))

where tv is the vector of net exports in value terms, ƟT is the matrix of total (direct plus indirect) factor cost shares, W is a diagonal matrix of factor returns, is the vector of factor endowments, and a superscript j denotes countries. The above equation predicts that the total factor content of trade in value terms (ƟT tjv) is a linear function of national (Wjj) and world factor endowment income \(\sum\limits_j {{W^j}{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\leftharpoonup}$}}\to {v} }^j}} \).

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 39.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 54.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. The expenditure shares are normalized to add up to 1.

    Google Scholar 

  2. As will be shown in Section 8.2.2, the industry’s skill intensity is proportional to the number of high- to low-skilled workers within the industry or the average industry wage.

    Google Scholar 

  3. We denote by total world’s exports the sum of exports for the countries in the sample.

    Google Scholar 

  4. This tests whether the adjusted likelihoods of two rival models are compatible, and it is equivalent to checking variance encompassing.

    Google Scholar 

  5. Both the Pearson and Spearman rank correlation coefficients between the minimum wage and the ratio of high- to low-skilled workers are positive and strongly statistically significant, when computed for a sub-sample of 37 countries for which data are available in 1988. The Pearson correlation coefficient is 0.716, while the Spearman ranking correlation coefficient is 0.724, both being statistically significant at least at the 1% confidence level.

    Google Scholar 

  6. This refers to groups 5–8 at the first-digit of the SITC Revision 2, except group 68.

    Google Scholar 

  7. The average wage refers to the manufacturing industries only.

    Google Scholar 

  8. The Pearson correlation coefficient is 0.8941, while the Spearman correlation coefficient is 0.895 for 41 observations.

    Google Scholar 

  9. For details, see Section 7.2, the part referring to the definition of the independent variables in cross-industry regression studies.

    Google Scholar 

  10. Harkness (1978) is an exception, as he uses factor cost shares to predict the trade pattern for the US.

    Google Scholar 

  11. Statistical significance is given at the 1% confidence level, when using a two-tailed test.

    Google Scholar 

  12. This refers to countries with statistically significance correlations.

    Google Scholar 

  13. Exceptions are Guatemala and Pakistan for low-skilled labor, and Guatemala, Kuwait, Peru, and the Philippines for high-skilled labor. Also, for skill intensity, all the rank correlations, except for Hong Kong, are statistically significant. See Section 8.3.3 below and Table 2 for details.

    Google Scholar 

  14. The Pearson correlation computed for low-skilled labor is larger than that of high-skilled labor in 71% of the cases, while the rank correlation is bigger in 79% of the cases.

    Google Scholar 

  15. There are twice as many developed countries with a statistically significant correlation in 1989 than in 1978.

    Google Scholar 

  16. There are twice as many developing countries with a statistically significant correlation in 1989 than in 1978.

    Google Scholar 

  17. The figure is 82% when we consider the significant correlations only.

    Google Scholar 

  18. The figure is 62% when we consider the significant correlations only.

    Google Scholar 

  19. Data on foreign net direct investment were available only for 40 countries in the sample.

    Google Scholar 

  20. The comparison is done for the case where GNP data are taken from the World Bank.

    Google Scholar 

  21. For details of the derivation, see Section 4.2.2.

    Google Scholar 

  22. For details on the construction of the missing data, see Section 8.2.1.

    Google Scholar 

  23. These are mainly countries for which factors other than those considered here may be relatively more relevant for their trade pattern, e.g. Australia, Finland, New Zealand and Norway.

    Google Scholar 

  24. Trefler (1995) defines relative factor endowments using our first definition. The abundance of a country is judged upon the number of factors in which the country is predicted, by its factor endowments, to be relatively abundant.

    Google Scholar 

  25. More than 80% of the correlations between trade and skill intensity have the expected sign.

    Google Scholar 

  26. When factor prices differ across countries, the order of factor intensities may be reversed between industries. The absence of factor intensity reversal ensures that the ranking of industries by factor intensities is the same everywhere.

    Google Scholar 

  27. Data availability differs across countries, the number of 4-digit manufacturing industries ranging from 21 and 52. The countries for which data are available are listed in Section 8.2.1.

    Google Scholar 

  28. Statistical significance is given at the 1% confidence level, when using a two-tailed test.

    Google Scholar 

  29. Data are available only for 20 industries for Kuwait, hence the results may not be reliable.

    Google Scholar 

  30. Here, only the results when capital endowments are computed in US dollars, using a PPP adjustment for the exchange rates, are reported.

    Google Scholar 

  31. The countries whose the estimated coefficients are not jointly significant at least for one specification are excluded from the comparison.

    Google Scholar 

  32. Advertising expenditures are reported in the input-output tables for the US at the same aggregation level as the input-output requirements. Hence, they are available at a much lower aggregation level than the mark-up data from Morrison (1990).

    Google Scholar 

  33. For the perfect competition model, there is an almost equal number of developing and developed countries with a statistically significant multiple correlation coefficient while, for the IRS model, the ratio is 2 to 1.

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Rights and permissions

Reprints and permissions

Copyright information

© 1999 Physica-Verlag Heidelberg

About this chapter

Cite this chapter

Keuschnigg, M. (1999). Empirical Analysis. In: Comparative Advantage in International Trade. Studies in Empirical Economics. Physica-Verlag HD. https://doi.org/10.1007/978-3-642-50212-5_9

Download citation

  • DOI: https://doi.org/10.1007/978-3-642-50212-5_9

  • Publisher Name: Physica-Verlag HD

  • Print ISBN: 978-3-642-50214-9

  • Online ISBN: 978-3-642-50212-5

  • eBook Packages: Springer Book Archive

Publish with us

Policies and ethics