I see this error all the time in economics and it's so basic it drives me spare. First of all, deficits as a %of GDP are not reliable indicators of whether a government is engaging in stimulus or austerity because the budget automatically adjusts to economic activity. 
Apparently even Nobel Prize winners don't recognise this simple fact. As GDP drops the budget deficit is going to increase (or surplus decrease) as tax returns are lower (these are themselves a function of income) and unemployment benefits increase.
The other error is a basic mathematical one in that GDP appears on both axes! So as GDP increases the deficit/GDP will decrease ceteris paribus all on its own. It takes monumental mathematical incompetence to say that the latter is causing the former.
Economists will happily point to this correlation even though it has no grounding, but will rightly recognise the endogeneity issues when you correlate interest rates and inflation.
