We examined the 28 OECD countries defined as "advanced" by the IMF between 1965 and 2010. Using regression analysis to control for the growth rates of the factors of production (physical capital, labor and human capital) and initial GDP, our results suggest that reducing the ratio of taxes or spending to GDP by five percentage points increases the growth rate of GDP per capita by 0.5 to 0.6 percentage points per year.
A broader sample of all "advanced" countries (again, as defined by the IMF) over the past 10 years seems to support these findings. Over this period, countries whose governments tax and spend less than 40% of GDP have grown more quickly than the big-government countries.
These differences in growth rates are important. Small differences in percentage growth rates roll up to huge differences in wealth generation over a number of years. If the differential of the last 10 years were to be constant over 25 years, then the economies of those countries with small governments would have more than doubled (an increase of 115%) while big government countries would have only seen growth of 64%.
Is this conclusive proof that cutting the size of government will always increase growth? Of course not. The accumulation and quality of other factors are also important. But this evidence shows that other things equal, countries with small governments and with small tax burdens grow faster.
Reason.com on government spending.
Another view on shrinkage, significant shrinkage: