I am currently undergoing a study on economic growth in 11 countries: Bangladesh, Egypt, Indonesia, Iran, Mexico, Philippines, Pakistan, South Korea, Turkey and Vietnam. I plan to use panel regression. I plan to follow an analysis such as Islam (1995) and Barro (1991) with an augmented Solow model with human capital.

As my human capital variable I am using the "Index of human capital per person" (based on years of schooling and returns to education), from the Penn-world tables. My question is, how would I go about finding the savings ratio $S_h$ i.e. the investment in human capital from the above index?


If you multiply the index by the population size (or the quantity used to compute the per person index), you get an index of the stock of human capital. For example, if for a given country the index is $h_t$ and population is $N_t$, then the stock is:

$$ H_t = h_t \times N_t $$

Now, by the definition a stock follows an accumulation rule, just as physical capital:

$$ H_{t+1} = H_t(1+d_t) + I_t $$

where $d_t$ is the "depreciation" of HK (e.g. because of death or skill obsolescence) and $I_t$ is gross investment in HK. Since you probably do not have data on depreciation (which I imagine is very difficult to compute), you can only calculate net investment, which is equivalent to net savings by definition:

$$ S^h_t = H_{t+1} - H_t $$

Finally, to compute the savings rate, you need to divide this by a measure of GDP. However, these need to be measured in the same units. Therefore, you need to transform $H_t$ or $S^h_t$ into monetary units. For this you need something like the average price of human capital for every period, which I have never seen, and seems not available in Penn-world tables.

| improve this answer | |
  • $\begingroup$ Why the downvote? $\endgroup$ – luchonacho Aug 3 '17 at 9:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.