This essay is the take-home examination of the Economics 502 - Macroeconomic Theory 2.1 (2017) taught by Professor Boldrin. It is going to explore the empirical truthfulness of the famous facts stated by Nicholas Kaldor in 1957, by comparing them with data about Italy in the period 1950-2014.
In the last 40 years all the developed economies experimented a decline in the real interest rate and a slowdown in labor productivity. These topics are currently debated in seminars and conferences, nonetheless many economics classes and many textbooks about economic growth (e.g. Barro & Sala-i-Martin (2004) p. 12 and Acemoglu (2009) p. 57) feel the necessity to state some facts that predict the exact opposite,
These are the so-called Kaldor facts and they are a still workhorse in macroeconomics. They were firstly stated in Kaldor (1957) and then repeated in Kaldor (1961). In the first paper Kaldor states that model of growth has to take in account “the remarkable historical constancies revealed by recent empirical investigations” such as:
The share of wages and the share profits has shown a remarkable constancy in developed economies. In particular, Kaldor citing Phelps Brown & Hart (1952)1 mention that the share of wages in UK was roughly 40% in the period 1840-1950 and in US was 60% until 1929 and then it increases to 69% (Kuznets (1952))
Capital equipment and output per worker (at constant prices) rise steadily and the trend is the same so that the capital-output ratio does not change over long period. In particular, he reports that UK has a ratio equal to 3.3 between 1870 and 1914 and 3 between 1914 and 1938 and the US has the same ratio in 1890 and in 1950, even though for both countries upward and downward trends
Given that share of profit over total income is constant and the capital over output is constant then the profit over capital is constant (“rate of profit earned on investment”). As usual he show the data for UK (between 9.5%-11.5% in the period 1870-1914)
These facts have given an empirical support to the theoretical framework that is based on the idea of Balanced Growth Path (BGP) that is the main way of thinking about growth economics2.
Actually , just reading carefully the facts and the footnotes doesn’t seem these facts are so “facts”: they are limited to two countries and some of them consider a very limited period, always excluding the Great depression- Moreover the concept of constancy is very vague. In other words, they don’t seem to be empirically very robust.
In this essay we are going to test explanatory power of these facts by considering the case of Italy. In order to achieve this object in the following section we are going to discuss the data used. Then, in section 4 we are going to discuss the empirical evidence about Kaldor facts in Italy in the period 1950 and 2014. In section 5, we are going to discuss in a broader sense the idea of using balanced growth path to model the Italian growth. In section 6, we are going to verify which one out the models discussed during the class best fit the pattern of growth. In section 7 after a brief review of literature we are going to propose a new model.
In order to test Kaldor data we relied on the Penn Word Table version 9.0 (Feenstra et al. (2015)). The data present the following advantages: 1) the length (from 1950 to 2014), 2) the presence of all the variable we need to conduct our analysis, 3) the possibility of choosing the different . However, it would be interesting the counter-check this data
Which data do we need to test Kaldor facts
Total output (also called Total income, GDP, Y) in real terms
Total Stock of capital (Gross/ net) (K)
Number of workers
Total amount of Wages
Profits
When other sources will be used it will be indicated3
The first fact states that the share of wages (\(wL\)) and the share of profit \(1-wL/Y\)4 are constant in developed countries, it doesn’t seem that true (see Figure 1). In more detail, until the beginning of 80s we can see some constancy5, but after is clear a decreasing trend where labor share from being slightly above 60% reaches figures slightly above 50%.
Share of labour compensation in GDP at current national prices. Source: PWT9, Feenstra et al. (2015)(black), Ameco (green)
The second fact states that capital equipment and output per worker (at constant prices) rise steadily, but also here the statement is not reflected into the data, at least not along all the series.
Log real GDP (blue) and real capital (red) at national prices per worker. Source: PWT9, Feenstra et al. (2015)
In fact, while in the period 1950-170 the growth is steady and the trend is indeed the same, in the period 1975-1995 both the variable decrease their growth and eventually after 1995 the output per worker it is completely flat while the capital per worker keeps increasing (see Figure 2 and Figure 3).
Growth rate output-side GDP (blue) and Growth rate capital (red). Source: PWT9, Feenstra et al. (2015)
0.7055707
In fact the capital output ratio is not constant
Capital stock at current PPPs (in million 2011 USD) divided by Output-side GDP at current PPPs (in million 2011 USD). Source: PWT9, Feenstra et al. (2015)
Share of profit in GDP (\(1-labor share\)) at current national prices over Capital output ratio. Source: PWT9, Feenstra et al. (2015)(black)
Let’s check now some other variables to establish whether a balanced growth path exist
Logarithm Real consumption at constant 2011 national prices (in million 2011 USD). divide by the number of persons engaged. Source: PWT9, Feenstra et al. (2015)
Growth rate Real consumption. Source: PWT9, Feenstra et al. (2015)
Acemoglu, D. (2009). Introduction to Modern Economic Growth. Princeton University Press. Retrieved from https://books.google.com/books?id=jOj4P3NMreMC
Barro, R. & Sala-i-Martin, X. (2004). Economic Growth. McGraw-Hill. Retrieved from https://books.google.com/books?id=jD3ASoSQJ-AC
Feenstra, R.C., Inklaar, R. & Timmer, M.P. (2015). The Next Generation of the Penn World Table. American Economic Review, 105, 3150–3182. Retrieved from http://dx.doi.org/10.1257/aer.20130954
Kaldor, N. (1957). A Model of Economic Growth. The Economic Journal, 67, 591–624. Retrieved from http://www.jstor.org/stable/2227704
Kaldor, N. (1961). Capital Accumulation and Economic Growth. 177–222. Retrieved from http://www.fep.up.pt/docentes/joao/material/macro2/Kaldor_1961.pdf http://dx.doi.org/10.1007/978-1-349-08452-4_10 www.fep.up.pt
Kalecki, M. (1938). The Determinants of Distribution of National Income. Econometrica, 6, 97–112. Retrieved from http://www.jstor.org/stable/1907142
Keynes, J.M. (1939). Relative Movements of Real Wages and Output. The Economic Journal, 49, 34–51. Retrieved from http://www.jstor.org/stable/2225182
Kuznets, S. (1952). Long-Term Changes in the National Income of the United States of America since 1870. Review of Income and Wealth, 2, 29—–241. Retrieved from http://dx.doi.org/10.1111/j.1475-4991.1952.tb01048.x
Phelps Brown, E.H. & Hart, P.E. (1952). The Share of Wages in National Income. The Economic Journal, 62, 253–277. Retrieved from http://www.jstor.org/stable/2227004
In turn Brown and Hart cite Keynes (1939) and Kalecki (1938)↩
About this point I found interesting this article and this article reading from Professor Vollrath’s blog.↩
The idea would be to te↩
Worth to notice that if we don’t assume perfect competition the share of profit it is different from the share of capital income (\(rK\)). To be compare with Istat that in the income approach divide gross domestic product at market prices in domestic compensation of employees, gross operating surplus and gross mixed income, taxes on production and imports and subsidies (-)↩
We compared the labor share provided by PWT with the AMECO in order to check a really strong constancy that can be due to some lack of data (to be checked) but also AMECO data lie within a small range (6%) around the mean (63%)↩