Two years ago, Treasury published in the budget papers an analysis of the impacts of fiscal stimulus. The graph purported to show that there was a positive relationship between the size of stimulus and the subsequent revisions or changes in economic growth forecasts.
Sinclair Davidson did a job on Treasury exposing this analysis as ‘data snooping’. Treasury had left out 8 countries for which comparable data was available from their analysis and when they were included the relationship between stimulus and growth disappeared.
Treasury later responded to the Senate for their error, as David Gruen said:
Professor Davidson is correct. Before publishing the results in the Budget, the regression result for the full sample of 19 countries was checked. Unfortunately … an error was made and the erroneous conclusion was drawn that the results for the restricted sample did not differ, to any material extent, from those for the full sample.
Unsurprisingly, their has been no update from Treasury on the data since.
Now, however, we have enough data to use the latest IMF World Economic Outlook to update the graph with the actual data from 2009 and 2010 (some of the data in Treasury’s analysis used forecasts). There basically remains no relationship between stimulus and growth according to Treasury’s test.
The relationship looks even worse for Keynesians if you include the data for 2011 (some of which remains forecasts for some countries).
(One limitation here is I am using the same data on stimulus that Treasury used, which is only for 2009 and 2010 and is forecasted stimulus not actual.)