Thursday, 1 January 2015

Growth in government spending and growth in GDP

Robert Waldmann shows that, in the US, quarterly growth in government spending has been positively related to quarterly growth in GDP since the third quarter of 2009. Yet Stephen Williamson shows that this relationship turns negative when just a few earlier quarters are added to the analysis. Importantly, both Waldmann and Williamson caution that not too much can or should be inferred from a bivariate analysis of ~20 time points.

Just for amusement's sake, I thought I'd examine the relationship between annual growth in government spending and annual growth in GDP for the UK, over all 58 years since 1956. Data are from the Office for Budget Responsibility; nominal GDP was adjusted for inflation using the GDP deflator. Interestingly, as the chart below indicates, the relationship is almost perfectly null. Rounded to two decimal places, the regression coefficient is 0.00, with a p-value of 0.98. When using the previous year's growth in government spending, the coefficient is –0.07, with a p-value of 0.45.

No comments:

Post a Comment