Skip to main content

Advertisement

Log in

Applying panel vector autoregression to institutions, human capital, and output

  • Published:
Empirical Economics Aims and scope Submit manuscript

Abstract

We bridge two areas of study by applying panel vector autoregression (PVAR) to human capital, political institutions, economic institutions, and economic output per capita. Institutions and human capital have competed within the scholarly literature as hypotheses explaining the origins of economic growth. Elsewhere, our measure of economic institutions, the Economic Freedom of the World index, has recently been explored extensively as a dependent variable, whereas previously it had been used as an explanatory variable. We wish to measure the interrelationships between political and economic institutions, as well as their interrelationships with economic output and human capital, in contrast to the literature which emphasizes the importance of political institutions alone. We explore these interrelationships in a PVAR model, finding that, descriptively at least, higher-quality economic institutions are associated with more output. We also find weak evidence that higher-quality political institutions are associated with less output and less education. We also find a robust positive effect of education on the quality of economic institutions. In performing this analysis, we contribute to the literature on the institutions and human capital debate, as well as to the literature on the causes of free economic institutions.

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

Access this article

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Notes

  1. Regarding the relationship between education and beliefs in effective economic policy, see Caplan (2001) and Caplan and Miller (2010). Regarding the interrelationships between economic freedom and political freedom, see Farr et al. (1998) and Dawson (2003).

  2. The Penn World Table only currently runs through 2014. For the last period in our data, 2015, we use the data from 2014 as a substitute.

  3. At least five time observations are required to be included in the “full sample.”

  4. Regression results are meaningless if the variables are of a different order of integration (Enders 2010: 199).

  5. Results from the Pesaran (2007) panel unit root test, which is robust to the presence of cross-sectional dependence, suggests that all four series contain a unit root.

  6. Harris and Tzavalis (1999) show that the assumption that T is asymptotic yields a test with reduced power short panels.

  7. Results available upon request.

  8. Prior to performing the Helmert transformation, each series is differenced to achieve stationarity. See Sect. 3 for the corresponding discussion and unit root tests.

  9. We avoid alternative estimators, such as system GMM, that take advantage of additional moment conditions but require stronger assumptions, specifically assumptions about initial conditions (Bun and Sarahdis 2015: 85).

  10. See Sect. 3 for more on the non-stationarity of the series.

  11. Blundell and Bond (1998) suggest the system GMM estimator, which takes advantage of additional moment conditions by making assumptions about the initial conditions.

  12. All series in the present study are differenced to ensure stationarity. After differencing, the estimated eigenvalues for each series are less than 0.5 in absolute value and therefore lie well within the unit circle.

  13. In fact, increases in the cross-sectional dimension may reduce bias in GMM estimation of a PVAR system as in Binder et al. (2005).

  14. 90% confidence is actually a higher standard than is often used in much of the other literature employing vector autoregression. See Murphy (2015).

  15. Impulse response functions available upon request.

References

Download references

Acknowledgements

The authors would like to thank Vasudeva Murthy for his helpful comments and suggestions.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Ryan H. Murphy.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Murphy, R.H., O’Reilly, C. Applying panel vector autoregression to institutions, human capital, and output. Empir Econ 57, 1633–1652 (2019). https://doi.org/10.1007/s00181-018-1562-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00181-018-1562-0

Keywords

JEL Classification

Navigation