Human Capital and National Security

Abstract

This study evaluates the relationship between a country’s human capital and its capacity to achieve, over time, effective national security. We show that an increase in the importance that a country attributes to its future, and/or in the effectiveness of its education system, leads to a decline in its current civilian services but to increases in its future human capital, civilian services, national security, and social welfare. Data for the Israeli-Iranian arms race confirm our predictions and show that a predilection for short-term political or personal gains impedes investment in human capital and inhibits future economic growth, national security and social welfare.

Keywords:

• Arms race
• budget allocation
• education
• human capital
• national security

JEL codes:

• D74
• D78
• D90
• H56
• H68

Notes

1. For example, the World Economic Forum's Global Human Capital Report (2016) ranks Israel 23rd with a score of 78.99 compared with Jordan’s 64.7 (81th place), Iran’s 64.16 (85th place) and Egypt’s 63.72 (86th place). In 2015, Israel’s education system received a score of 0.870 in comparison to Iran’s 0.704, Jordan’s 0.701, Egypt’s 0.601 and Syria’s 0.417. The Bloomberg Innovation Index (2017) ranks Israel 10th, overall and 2nd in the R&D index. None of Israel’s close neighbors appears in this ranking.

2. The magnitude of the effect of human capital on economic growth depends, among other things, on institutional factors. Rogers (2008) shows that the effect of schooling on economic growth depends on the country’s characteristics, such as corruption level and the extent of brain-drain that it experiences.

3. On the recent trends in military R&D see Hartley (2006) and Mowery (2012).

4. National Defense Budget Estimates for FY 2015.

5. For a comprehensive discussion on the response to threats and use of force, see Powell (1999).

6. See Smith, Sola, and Spagnolo (2000) on the Greek-Turkish conflict, (Snyder 1971) on international conflicts, and Plous (1993) on the nuclear arms race.

7. See Kagan, Tishler, and Weiss (2005) on the response of developed countries to terror weapons and Kagan et al. (2009) on the characteristics of the arms race between South- and North Korea.

8. Wan (2012) analyzes the relations between the quality of military personnel and the effectiveness of the military. His study serves as a reference for policy makers on the use of military recruiting systems and other military manpower policy levers used ‘… to attract, recruit, train, retain, maintain and enhance the military human performance’ (183). Additional assessments of the relationship between human capital and military effectiveness appear, among others, in Nolan (1993), Asch, Can, and Schonlau (2004), Asch, Romley, and Totten (2005), Williams (2005), and Asch, Hosek, and Warner (2007).

9. It is common to use static arms race models. Modeling and solving dynamic arms races is a difficult analytical challenge. Nevertheless, arms races may be dynamic by their nature (see Intriligator 1975; Richardson 1960).

10. The country’s aggregate production can be interpreted as its GDP.

11. Aggregate production as a function of human capital stock is employed, for example, by Acemoglu (2009), Glomm and Ravikumar (2001) and Gradstein and Justman (2000).

12. See Viaene and Zilcha (2002, 2013) for similar formulation of a human capital production function.

13. We assume that the military power of a country depends on its GDP and arms procurement, relative to those of its rival, but not directly on its human capital stock. The quality of weapon systems is embedded in the price of these systems and, hence, it is represented in the variable arms procurement. Efficient use of higher quality weapon system may require higher human capital. For simplicity, we do not include human capital in the specification of the relative military power function (Equation 3). However, the stock of human capital is directly related to aggregate production, which represents in Equation (3) the value of human capital, economic prowess and, possibly, other factors (which are captured by the parameters α ⁱ and A ⁱ ).

14. See Hirshleifer (2000) for a discussion of various measures of military capability and national security.

15. Using partial depreciation of weapon systems over time precludes the derivation of an analytical solution of the model, but does not change the nature of its solution. Hence, we assume full depreciation of weapon systems in our model.

16. The definition of the perception of national security is adapted from Bar-El, Kagan, and Tishler (2010) and its interpretation is similar to that in Levine and Smith (1995)), Golde and Tishler (2004) and Shabtay and Tishler (2014). The perception of national security by each country is increasing, albeit at a decreasing rate, in the country's aggregate production and its expenditure on weapon systems.

17. Analytical tractability requires the simplifying assumption that the arms stock fully depreciates at the end of each period. Using partial depreciation of weapon systems over time does not change the nature of the results in the paper, but precludes the derivation of analytical results.

18. We also assessed arms races between Israel and Egypt, and Israel and a coalition which includes Iran, Egypt and Syria using publicly available data for 1993–2013. The policy implications of these arms races are similar to those presented here and, hence, are not reported here.

19. For the last two decades, Iran has repeatedly threatened publicly to destroy Israel and, therefore, is considered to be Israel’s main arms race rival (see, for example, Haaretz 2015; Sadjadpour 2018; Sadr 2005). For simplicity, we ignore the possibility that one of the rivals, or both, may face other potential arms race rivals.

20. The theoretical model provides some clear cut conclusions about the effects of a change in the model’s parameters or exogenous variables on the model’s endogenous variables. The role of the calibration process and the policy experiments is to assess the magnitude of such changes. A cost benefit analysis may be required to assess whether a particular policy, which is costly to perform, is desirable to the country. Such analysis requires to estimate the model’s parameters and compute, often in monetary terms, the benefits and costs of the policy. Note also that the effects of some changes in the model’s parameters cannot be deducted from its theoretical structure, but depend on the size of the model’s parameters (and exogenous variables). For example, it is not straightforward to predict how a change in importance that a country attributes to its future affects its perception of security. Furthermore, in an arms race setup, a change in a parameter may affect both countries endogenous variables, which, in turn, may affect each other. The net effect of such a change, and certainly its magnitude, may not always be predicted from the theoretical model, but can be evaluated for a particular arms race (a set of calibrated parameters and exogenous variables). Finally, the speed and curvature of a change of an endogenous variable over time, in response to a one-time (or permanent) change in a particular parameter (exogenous variable) may not always be deduced from the theoretical model, and its assessment may require calibration and policy experimentation.

21. See Amemiya (1985) and Greene (2012) for the properties of various estimation methods of linear and nonlinear systems of equations. We employ single equation estimation method, such as OLS or 2SLS, when the estimation is carried out for a single equation of each country, and a system of equations estimation method, such as nonlinear 3SLS (N3SLS), when we estimated the parameters of a particular equation simultaneously for both countries.

22. The stock of human capital is given by an index, hence, its units of measurement are arbitrary. Such updating procedure is common in economics (see Setter and Tishler 2006; The World Economic Forum 2016).

23. We employed the country's overall human capital index as a proxy for the effectiveness of its education system. That is, $B^k=1+\frac{{\left(overallindex\right)}^k}{100},fork=i,j$. The values of the overall index were taken from The Global Human Capital Report 2015.

24. We do not have explicit data for human capital. Thus, to limit the effect of the initial condition (the initial value of the human capital data series), in constructing this series we employed data from 1989 to 2013. The calibration procedure employed data for 1993–2013.

25. The parameters were estimated by the statistical package EViews after taking the logarithms of both sides of Equation (1). GDP was used as an approximation to aggregate production.

26. For details on the use and properties of N3SLS see Amemiya (1985).

27. The choice of values for $b^i$ and $b^j$ (the anxiety effects) sets the units of measurement of the national security levels for the two countries. That is, this choice is arbitrary.

28. The parameters were estimated by the statistical package EViews.

29. Similar changes in Iran’s parameters yield similar conclusions.

30. It is impossible to provide reliable long-term predictions about the future path of the Israeli-Iranian arms race. That is, the projections in this study are presented in order to clarify the properties of the dynamic arms race model.

31. This kind of shortsighted policy is not confined to right wing parties; on the contrary, left wing parties in western democracies seem to extend military expenditure as part of their welfare program (Eichenberg and Stoll 2003; Whitten and Williams 2011).

32. Preference for short-term policy over a longer one may also be the result of sticking to an ‘inferior’ Nash equilibrium (due, say, to the prisoner’s dilemma problem), or the inability of governments to plan and execute complex long-term programs.