Are There Laws of Production? History

Are There Laws of Production? HistoryThe cost of labour, if we disregard periods of economic hardship, has a clear tendency for growth. Given the continuous quantitative increase in capital, confirming whether over time the relation between the cost of labour and the cost of capital in a certain country or industry changes or not becomes difficult. Will the fact of equalization of the rate of profit in various industries not serve as confirmation of the similar relationship between the share of labour and capital in the final value of a product in all sectors of the economy? This is all very interesting, but here we must stop and continue the discussion of Douglas’ results.
At the end of the 1920’s, using data from American manufacturing industry from 1899 to 1922, Paul H. Douglas attempted to determine the value of the labour exponent k using the method of least squares. It turned out to be 0.75. It was established that the discrepancy between the theoretical value of the product, which was determined using the production function and the real value was small. The most significant variations were observed in the years of the depression and maximum economic activity. This was explained simply by the incomplete nature of the indexes of labour and capital. The indexes measured the number of factors of production rather than the degree of their relative use. Evidently, during the years of depression the usage level of equipment was low and employment was part-time. During the years of prosperity, equipment was used extremely intensively and employees were hired to work overtime. Therefore these discrepancies were more likely an additional confirmation of the overall value of the formula for normal periods. Further confirmation was received from income analysis performed by the National Bureau of Economic Research. It was established that the share of labour in the net value of an industrial product from 1909-1918 was 74.1%, i.e. almost exactly corresponded to the value of the exponent for labour.
These results looked promising and Cobb very quickly joined in the analysis. He calculated the indexes of labour and capital in manufacturing industry in Massachusetts for the period from 1890 to 1926 and found that the value of k was 0.743. A similar study carried out for manufacturing industry in New South Wales for the period from 1901 to 1927 gave a value of k=0.65. In 1937, data was analysed for the industry in Victoria for the period from 1907-1927 and the k value was 0.71. After the war, studies were carried out for American manufacturing industry in the following years: 1889, 1899, 1904, 1909, 1914 and 1919, for Canada in 1923, 1927, 1935 and 1937, three studies for Victoria in 1910-1911, 1923-1924, 1927-1928, one study for New South Wales in 1933-1934, five studies for Australia in 1912, 1922-1923, 1926-1927, 1934-1935, 1937-1938. Douglas’s students added to the studies with data for Queensland for 1937-1938 and New Zealand for 1926-1927. Structural analysis was also carried out for South Africa. As a result the following values for the exponents were obtained. For American manufacturing industry the k value was in the range of 0.63-0.64 and j – 0.34, for Australia the average value of k was 0.60 and j – 0.37, for South Africa in 1937-1938 k was 0.66 and j – 0.32. The data for Canada and New Zealand were slightly different: the value of the exponent of labour was lower and the exponent of capital was higher (for Canada – k = 0.47, j =0.52, for New Zealand – k = 0.42, j =0.49).
What conclusions do the results presented allow us to make? There is a surprising constancy in the share of labour and capital throughout the period of observations in each of the countries studied. For the US, Australia and South Africa the share of labour in manufacturing industry was close to 2/3 and the share of capital – close to 1/3. For Canada and New Zealand there were slightly different results. This can undoubtedly be linked to the particular features of recording and gathering statistical data in the different countries. This may also reflect differences and peculiarities of development of economies in these countries. However, the idea seems highly interesting. With an increase in capital by 1% in countries such as the US, Australia and South Africa, the product increases by 1/3% and with an increase in the number of hired workers by 1%, production increases by 2/3%. For a balanced and optimum expansion of production, the amount of capital and labour must be increased in the ratio of 1 to 2. In this case production will be as effective as possible and there will be no reduction in output: with an increase in costs by 1% production will increase by 1%. These conclusions seem very interesting and as noted by P.H. Douglas, they cannot be called a mere coincidence. The results obtained must without doubt be explained by a certain important dependency, or, in general, by a certain law of production. But what does it consist in and what it is its essence? Why, after spending a certain sum of money on acquiring additional capital, should a manufacturer spend exactly twice as much on hiring additional labour? What mechanism forces the manufacturer to do so? The answers to these questions were not found. We would like to highlight another, now widely-known fact that involves the strange constancy in the share of consumption in the GDP. (Reder, 1959; McConnell and Brue, 2005). That share is 2/3. To consider this a coincidence would be entirely unreasonable. Let us try to understand all of this and provide a plausible explanation for these facts. To start off with, however, we need to discuss the nature of savings and investments, since it is they that are responsible for the generation and arrival of additional capital. And the ratio between consumption and savings determines the share of labour and capital in the end product.