This paper attempts to:

- compare GDP per capita level and growth across 17 advanced countries over 1890-2013.
- compare the level and growth of the main components (TFP, capital intensity, working time, and employment rate) of GDP per capita in order to see how they contribute to the GDP per capita difference.
- test the convergence hypothesis of GDP per capita and its components over different sub-periods.

The second half of this paper focuses on convergence. This is the hypothesis that economies with lower per capita income will tend to grow faster than the ones with higher per capita income. There are mainly 3 types of convergence: absolute convergence, conditional convergence, and club convergence. One of the two approaches to convergence is sigma-convergence. It refers to the reduction of dispersion of levels of income across economies with time.

Through data presentation and analysis, it yields the following results:

- All countries have at least one huge growth in GDP per capita in the 20
^{th}century, but in a staggered manner. - Almost all countries have faced a huge decline in GDP per capita growth.
- The GDP per capita leadership shifted over years.
- Overall convergence among advanced countries.
- GDP per capita convergence to the leadership position is not always happening.
- Employment rates and hours worked did not contribute to the overall convergence process.

My observation:

The paper claims that its analysis’ originality is that “it is presented over a long period, on a large set of countries, with data reconstituted in purchasing power parity and on the basis of, as much as possible, consistent assumptions.” However, most of the results it yields are simple observations such as “all countries experienced at least one big wave of GDP per capita growth during the 20

^{th}Century, but in a staggered manner” and “all most all countries have suffered, during the last decades of the period, from a huge decline in GDP per capita growth.” I think the paper could develop more into the implications of these observations. Also, since this is a time-series analysis, endogeneity and dual causality might present and I think the paper should address a bit more on that. For example, the paper talks about the impact of institutions on the components of GDP per capita. Is it possible that those components shape the institutions as well?
## 3 Comments

Indeed, the paper doesn’t break new ground; it’s good to spot that. It does go through several different criteria for multiple countries, and does present both long-run annual data and for recent years quarterly data. So if you’ve never seen a convergence paper before, it does give you a flavor of what is possible, and (less clearly) what we can’t tell from the data.

The graphs on contributions to growth are impossible to read, you can’t distinguish which is which with the grayscale “color” scheme. My hunch is that there are no clear results or the authors would have presented them more clearly.

The groupings are also hard to figure out, and if I read the tables correctly, a modest change in the time period leads to different groupings.

Are there any questions that they might be able to ask, using their dataset, but did not?

The paper doesn’t seem to produce many specific results, with the first result just being each country experienced a wave of growth. More could have been developed out of policy implications rather than concluding that policy “may influence relative GDP per capita levels,” which isn’t exactly conclusive.

Ay, a problem typical of growth regressions – the number of factors exceeds the number of observations, while (reasonably) lots of things matter. Pinning down details is decidedly difficult.

Comments are closed.