The short answer is “no”. Regardless of how complicated I make any regression analysis on measuring some ethereal polling bounce for the government caused by the budget, I simply cannot find one.
To highlight why this is the case, let’s look at the Newspoll results going back to 1990 and rather than bore the bejesus out of you with acres of regression results, we’ll just compare two measures of the polls.
Firstly, we’ll look at the two poll primary vote average of the government both before and after the budget, and take the difference between those results (to measure any short term change), and secondly we’ll do the same with a 4 poll average to measure a medium term change. For this, we’ll use Newspoll data.
Each budget essentially needs to be treated on its own merits. Some budgets, like Keating’s notorious 1993 one, turned out to be catastrophic in terms of its impact on both short and medium term levels of political support. Others, like Howard’s Porkstorm election budget of 2001, caused an increase in both short and medium term levels of political support for the government.
We often hear a lot of piffle from columnists around budget time, over the inevitability of budget bounces and “masterclass” budgets shifting political heaven and earth.
So it’s worth keeping in one’s thought orbit that, on average, budgets struggle to make a rats arse of difference – the occasional one does, but most simply do not. There is certainly no evidence to remotely suggest that budget bounces are a statistically significant phenomenon of Australian federal politics.