The Facts Are In, More Government Means Less Growth
Governments around the world have tried a myriad of policies in mostly fruitless attempts to help their economies recover from the recent severe recession. These policies, both tried-and-true and never-seen-before, have sparked a spirited debate among economists, central bankers, and policymakers about the role of government in economic policy. In particular, possibilities include that governments do little, cut spending, increase spending, cut taxes, increase spending a lot, print money, lower interest rates, or increase spending a lot and cut taxes. All of these have been tried in different countries around the world. A look at the data gives us an early indication of what does not work: government spending.
It is still too soon for a definitive answer on the efficacy of the various policies tried in different countries. Further, given that other factors were at work besides a large recession and financial crises felt in most parts of the world and the variation in conditions from country to country, it will take some sophisticated economic modeling to sort everything out in the end. However, we do have enough data to take a quick look and make one early conclusion.
To do so, I collected data from the World Bank on every country for which it had the data on government spending and GDP from 2004 to 2011. I ended up with complete data for 104 countries ranging across six continents and from rich to poor. I collected data on each country's GDP, its government spending as a percent of GDP, its government spending in dollars per capita, and its government spending in dollars.
The idea was to measure government spending three different ways: its share of the economy, its actual level, and the amount spent per person. I did this just in case the manner in which one determines whether government spending is "big" or "small" mattered to the result. In the end, it did not.
I used the GDP data to compute the growth in GDP from 2004 to 2011. This period includes a few years before the recession, the entire recession, and the first two years of the recovery. I would have used more recent data in order to include more of the recovery, but the government spending data for 2012 were very incomplete.
Next, I computed the correlation between each of the three measures of government spending (share, total, and per capita) and the growth in GDP over the seven year span. A correlation coefficient measures the tendency of two variables to move together. If one tends to go up when the other goes up, the correlation will be positive; if they move in opposite directions most of the time, the correlation will be negative. A correlation of zero implies the two variables are unrelated, positive or negative one implies an exact relationship between the two.
So if more government spending is good, the correlations should be positive. If they are negative, government spending leads to slower (or negative) economic growth.
The results showed that all three correlation coefficients were negative. The correlation between GDP growth and government spending's share of GDP was -0.44, pretty strong and very statistically significant (for those who can remember their college statistics class at least vaguely). That means economies with larger government sectors grew more slowly over that seven year period.
The correlation between GDP growth and per capita government spending was also -0.44. The correlation between GDP growth and the dollar amount of government spending was -0.25. Thus, no matter how I looked at it, more government spending means less economic growth.
These data do not reflect whether a country was running a government budget surplus or deficit during this period (although nearly all of them, especially the developed ones, were running large deficits), so it does not directly address the value of running a deficit in the attempt to gain a Keynesian economic stimulus.
However, the level of government spending is generally under the government's direct control, while tax revenues and, therefore, the deficit are only indirectly controlled by government. Also, it is government spending and the programs that spend all that money that tend to create the largest distortions in the free market. Taxes do cause some distortions, but in most cases these are smaller than those caused by government spending. Thus, focusing on spending makes sense.
Politicians have an obvious interest in larger government. More spending means more power, through the ability to direct the spending to the politicians' chosen priorities and often supporters. Regardless of the impact of government spending on economic growth, government will inherently want to be large and to spend as much money as possible. However, this look at data from around the world suggests that government spending comes with a cost. If a country wants to get richer, it should reduce its government spending.