Professor Costas Milas, Professor of Finance and Michael Ellington (NWDTC-ESRC PhD student):
“Investment uncertainty is a valuable missing piece of the productivity “puzzle” in the UK economy. Political uncertainty regarding the forthcoming general elections as well as a possible BREXIT referendum have the potential of adding to investment uncertainty and therefore delaying a steady recovery in productivity.
The ongoing productivity “puzzle” (very weak productivity growth in the aftermath of the 2008-2009 financial crisis) leaves little room for strong wage increases.
Consequently, tax revenues are weaker than otherwise leading to a persistent budget deficit. Much of the discussion about the productivity “puzzle” focuses almost exclusively on the relationship between productivity growth and real wage growth.
This, however, ignores the potential impact of investment uncertainty. In general, periods of economic stress are associated with elevated uncertainty. Consequently, firms pause hiring and investment.
In the latter case, productivity growth takes a hit because the drop in hiring and investment reduces the rate of reallocation from low to high productivity firms; the negative effect also (arguably) transfers to wage growth.
Using the Bank of England’s historical dataset, we plot together annual productivity (per head) growth, real wage growth and our measure of investment uncertainty. Productivity growth and real wage growth are 5-year moving averages.
We construct investment uncertainty by taking the 5-year moving standard deviation of gross fixed capital formation growth (see Data Appendix).
Figure 1 plots together productivity (per head) growth, real wage growth and investment uncertainty over 1861-2013.
From Figure 2, which zooms into the 1997-2013 period, we see that low investment uncertainty coincides with steady productivity growth and real wage growth. Nevertheless, elevated uncertainty during the (recent) financial crisis in 2008-2009, takes a hit both on productivity and real wage growth.
Figure 3 plots the posterior median (together with the 16th and 84th percentiles; dotted lines) response of productivity growth and real wage growth to a 1-standard deviation investment uncertainty shock in 2008.
From Figure 3, investment uncertainty exerts a negative and statistically significant impact on productivity growth and real wage growth of up to 5 years. Our calculations are done within a 3-variate Time-Varying Parameter Bayesian Vector Autoregressive model (using a Minnesota prior) which is conditional on exogenous World War I and World War II effects.
What are the implications of Figure 3?
- Our model goes some way towards explaining why productivity and real wages have not fully recovered yet following from the big investment uncertainty shock during the financial crisis of the late 2000s. We therefore argue that investment uncertainty is a valuable missing piece of the productivity “puzzle” and has to be considered by econometric models.
- From a policy point of view, political uncertainty regarding the forthcoming general elections as well as a possible BREXIT referendum have the potential of adding to investment uncertainty and therefore delaying a steady recovery in productivity.
Data Appendix
Labour productivity per head: We then construct annual growth and its 5-year moving average.
Wage index (spliced) and Consumer price index-original measure. We deflate the wage index by the consumer price index. We then construct the annual real wage growth and its 5-year moving average.
Gross fixed capital formation, composite measure, volume. We then constructed the annual growth and its 5-year moving standard deviation. This is our measure of investment uncertainty. All variables come from “three centuries of data”.