Wednesday, May 06, 2020

Are current productivity measurements still meaningful in services dominated economies? - a note

Back on 22 April 2020, Winton Bates asked What are the implications of declining productivity growth in high-income countries? Winton had previously expressed some reservations about statistical measures that suggested a structural decline in productivity growth in developed economies, in part because of difficulties in measuring the impact of information and communications technologies including real price declines. Now more convinced, he looks at the implications for future income levels.

I am very out of touch now with productivity discussions, but felt that the rise of services created difficulties for productivity measurement and indeed for the measurement of real income growth. I saw three problems.

The first lay in the definition of capital investment itself where measurement based on physical capital underestimated real capital investment in services where investment in things such as improved processes had a high labour component, a low physical component. A second problem lay in valuing outputs especially in non-market areas. A third and related problem lay in the assessment of quality in outputs.

My feeling was that, in combination, these problems meant that the level of capital investment in services was underestimated, output was underestimated, leading to misleading conclusions about the nature and level of productivity growth in services. Overall, I thought that productivity growth in services had been considerably underestimated.

I still think that was right, but now I suspect that productivity growth in services has declined, in part because the previous gains from process improvement have been largely exhausted while the longer term costs from process improvement have become more apparent. Services have also suffered from a growing regulatory burden, especially but not just where Government funding is involved. I wonder, too, about the quality of services. Has this actually declined as I suspect. 

These are not insignificant issues now that services constitute such a significant portion of the economies in developed countries, some 64%. I wonder how much meaning conventional productivity measurements now have.   


No comments: