Bryan Caplan took me to task for trying to separate size of government from other interventionist policies:
The truth is that size of government and economic freedom are inextricably connected. Any definition of “economic freedom” that doesn’t directlyincorporate the size of government is a crummy definition.
To illustrate, consider the following hypothetical. The government of Ruritania allows consenting adults to sell one another anything on any mutually agreeable terms. Ruritania has no minimum wage restrictions, no hiring or firing restrictions, no licensing, no zoning, and no paternalism. It even – wonder of wonders – has totally open borders. Anyone can hire anyone regardless of their national origin.
Before you packs your bags, I should point out that the government of Ruritania does have one little function. Namely: It imposes a 100% tax rate on all income, and redistributes that income equally to all. To enforce this tax rate, Ruritania has an all-pervasive system of surveillance – and punishes tax evasion with torturous death. Leaving the country counts as tax evasion.
By Scott’s standards, Ruritania is a free-market utopia. But almost no one else – economists, non-economists, or its own citizens – would see it that way. Ruritania is functionally equivalent to North Korea.
OK, let’s do another example. Start with a country like Australia, where government is 36% of GDP. Then pass a law stating that henceforth each person must pay, in addition to their regular taxes, another lump sum tax equal to 64% of gross income. However each person is also entitled to a welfare benefit equal to 64% of gross income. Voila, you’ve just raised taxes and spending to 100% of GDP. And yet of course absolutely nothing has changed. Now of course no country with an economic model worth considering is anything like my example, or like Bryan’s example.
But what this thought experiment does show is that spending/taxes as a share of GDP may be a highly misleading indicator of the degree of government intrusiveness. So what separates Bryan’s example from mine? It turns out that we actually care about other factors, which are very imperfectly proxied by spending ratios. For example:
1. High MTRs on consumption/labor discourage work.
2. Taxes on saving discourage investment and are biased against future consumption.
3. Taxes on estates discourage saving and investment.
4. Tax codes filled with loopholes encourage wasteful tax avoidance, and lead to wasteful expenditures of time and money on tax preparation.
Now let’s compare a country like Sweden to the US. Their total tax bill is much higher, but that tells us very little about how much we are interfering in the free market. Sweden has much higher taxes on labor/consumption (bad) but no marriage penalty (good). They have no inheritance tax (good) and similar MTRs (I think) on capital income. And they have a very simple income tax system (no forms to fill out), they just send you the bill (very good.) How does all this net out? I’m not sure. But what I do know is that we should pay much less attention to the fact that spending is 52.5% of GDP in Sweden and 38.9% of GDP here, and much more to how taxes affect one’s life.
I’m certainly not claiming Sweden’s system is “better” in any overall sense. Rather that the traditional “capitalist” and “socialist” labels don’t really capture the actual differences between the two economies. I will continue to consider it useful to differentiate between statist polices and egalitarian policies, even though I accept Bryan’s claim that virtually all real world egalitarian policies are at least slightly statist.
However I can’t accept this argument:
The essence of a free market isn’t merely that people can buy and sell whatever they want on whatever terms they find mutually agreeable. Without the right to keep what you earn, freedom of contract is utterly hollow. A society that redistributes most of what you earn is economically unfree.
. . .
Swedish school vouchers sound great. Nevertheless, Sweden’s tax/GDP ratio in 2011 was almost double that of the U.S – 47.9% versus 26.9%. The comparison for spending/GDP isn’t quite as lop-sided, but it’s still pronounced: 52.5% versus 38.9%. If you don’t think that weighs heavily against Sweden’s free market credentials, you’re making a big mistake – and abusing language along the way.
I reject the term ‘heavily.’ As I tried to show above, government spending and taxes as a share of GDP can be highly misleading. It doesn’t tell us how much “redistribution” is occurring, for instance. To a large extent the Nordic social welfare programs take money from the middle class and then return the money to the middle class. In contrast, Singapore forces people to take care of their own medical/retirement, etc. So the “official” level of government spending in Singapore is very low. And I do prefer the Singapore approach, which involves less redistribution and lower MTRs. But the G/GDP ratios vastly overstate the difference between the Nordic and Singaporean models. That’s why I focus on variables like MTRs and tax complexity, which do a far better job of picking up the extent of market distortion.
BTW, I was pleased to see Garett Jones raving about Prescott’s insights into tax theory, particularly the fact one should generally ignore the income effect when considering the effect of taxes on labor supply, and that a consumption tax is identical to a payroll tax. We learned all that at Chicago in the 1970s—glad to hear the rest of the world is now catching up.