Is There a Hauser’s Law for Canada Too?

            Economics has a number of “Laws” floating around that are rooted in empirical observation and then put forth as natural inexorable systemic laws.  For example, in public finance, there is Wagner’s Law of Expanding State Activity, which links the size of the public sector to income.  In health economics, there is Roemer’s Law, which basically states that a hospital bed built is a bed filled – a supply creates its own demand statement for health care.  And most recently, there has been popular reference to what is known as Hauser’s Law in response to the U.S. debate over resolving its fiscal deficit. 

 

            Hauser’s Law was formulated by W. Kurt Hauser of the Hoover Institute about 20 years ago and basically states that no matter how aggressive the tax regime, federal tax revenues as a share of GDP in the United States since 1930 have averaged just under 19 percent.  The implication is if the revenue share of GDP is fixed, then by lowering taxes and stimulating economic growth, one would get a fixed share of a larger economic pie thereby raising revenues.  The extension to all this is that in the face of the U.S. fiscal deficit debate, raising taxes would simply reduce federal revenues and make the deficit situation worse.  You would be better off according to the proponents of this view to cut spending and lower tax rates.  This is in essence a supply side type argument.  Of course, if the share is fixed, one could also make a case to cut tax rate rates and increase spending on stimulus programs, as the rise in total revenues from the stimulated economy should theoretically be able to cover much of the cost of the increased spending though I have yet to see this application of the logic in Hauser’s Law.

 

            In my younger undergrad days, the thought of inexorable laws that could explain economic phenomena really appealed to me.   I read Asimov’s Foundation series and as I learned about these laws they seemed like psychohistory come to life.  As time went on, I realized that while these laws may document some types of empirical regularity, one really needed a stronger theoretical base to motivate the law as well as a longer time span for the data.   In the case of Hauser’s Law, the question that comes to mind is why the federal revenue to GDP ratio is under 19 percent after the 1930s.  What was it before 1930?  I know it was substantially less in the pre welfare state/Keynesian era but could that not also be formulated into a "law"?  Moreover, I thought it might be interesting to see what the empirical regularities in the federal government revenue to GDP ratio are over time not only in the United States but also in Canada. 

 

            So, down below is a graph plotting total federal government revenue to GDP for both Canada and the United States over the period 1870 to 2009.  Over the entire period 1870 to 2009, the average total federal revenue to GDP ratio is 10.7 percent for the United States and 11.5 percent for Canada.  For the period 1930 to 2009, the ratio is 16.1 percent for the United States and 15.7 percent for Canada.  For the period 1870 to 1930, the ratio of federal government revenue to GDP was 3.6 percent for the United States and 5.9 percent for Canada.  Not surprisingly, the Great Depression and both world wars seems to increase the role of the state in the economy and the federal revenue to GDP ratio rises dramatically particularly during WW2.  Canada’s federal revenue to GDP ratio is in tandem with that of the United States but it is somewhat larger than the US before 1930 and somewhat smaller since 1930. In other words, Hauser's Law can also represent Canada during the post WW2 era.

Note that the federal revenue to GDP ratios reach a peak in the post-war period in 2000 at 20.4 percent for the United States and 18.2 percent for Canada.  They have since declined and in 2009 were at 14.8 percent for the United States and 14 percent for Canada.  Also, this decline is evident before the Great Recession, as by 2006 the revenue to GDP ratio had reached 18.0 percent for the United States and 15.9 percent for Canada. 

 

The structural economic shift brought about by Keynesian economics and the advent of the welfare state certainly can account for the increase in the federal revenue to GDP ratios since the 1930s.  In pre-Keynesian/welfarist state/interventionist government days, the federal revenue to GDP ratio fluctuated between 5 and 10 percent in Canada.  From 1870 to WW1, the ratio declines in the United States to well below 5 percent and then after WW1 is also in the 5 to 10 percent range until WW2.  Since WW2, the ratio for both countries has oscillated in the 15 to 20 percent range with the last decade seeing it approach the floor of the range. However, it has approached this floor before - in the late 1960s in both Canada and the United States and in the late 1970s in Canada - and then rebounded. 

So what next?  Will the federal revenue to GDP ratios in Canada and the U.S continue to oscillate between 15 and 20 percent or will the next 60 years see another structural shift in the role of federal governments in Canada and the United States that might generate a new range?  Based on the period 1930 to 1945, such a shift can happen relatively quickly. Unless there is some type of dismantling of the federal role in either or both countries, one should expect the ratio to stay in the 15 to 20 percent range.  Given the current political and economic climate, I certainly do not see it ratcheting upwards into the 25 to 30 percent range.

Comments

[Michael Reddell]

Interesting to see the long-term chart and how similar the Canadian and US federal experience is. But also interesting to think about the difference it makes to look at the issue at a general government level. At least over the last 20 years, general government outlays in Canada have averaged 43% of GDP, while in the US the average has been 37% (Source: OECD Outlook database). Any thoughts?

[Livio Di Matteo]

Michael:
Without looking at any more data, the difference might be attributable to greater outlays at the provincial-local level in Canada versus the state-local level in the United States.

[Determinant]

Is the one big budget difference between the US and Canada the provision of universal single-payer health care in Canada? Most every other program has a counterpart except that one.

Vermont, interestingly, has just passed a single-payer health law. They are going to implement a Canadian-style system in full by 2017.

It will be interesting to see what Vermont's tax rates end up as and provides an good test for Hauser's Law.

[Ironman]

Nice work, Liveo! If it's indeed like how Hauser's Law appears to work in the U.S., you'll find that the variation of Canada's tax collections as a percent share of GDP from year to year may be described by a normal distribution. That will give you the likely range into which the Canadian tax collections share of annual GDP will be observed for given levels of probability, say 68.2% of the time (+ and - one standard deviation) or 99.7% of the time (+ and - three standard deviations).

[Mike Moffatt]

"Is the one big budget difference between the US and Canada the provision of universal single-payer health care in Canada? "

Not really, no. Health care spending is roughly equivalent in the two countries thanks to U.S. Medicare.

[Mike Moffatt]

And by 'health care spending', I mean "government health care spending".

[Livio Di Matteo]

Ironman:
When you take the the percent changes in the tax revenue to GDP ratio and plot them, you do get a normal looking distribution. Here are the basic stats.

Federal Revenue to GDP Ratio (%)

Variable Obs Mean Std. Dev. Min Max

year 140 1939.5 40.5586 1870 2009
usa 140 10.7221 7.11996 1.79 23.85
canada 140 11.5059 5.62957 3.94 25.35
usa pct chng 139 1.76608 17.8845 -30.833 160.487
canada pct chng 139 1.53528 11.92244 -35.038 55.106

[Leo]

It seems to me that to change the ratio, all important governments have to make the change at the same time. The post-Great Depression change was only possible because everyone made it around the same time.

I think what's needed to change the revenue to GDP ratio is an international treaty to simultaneously change tax rates worldwide, so that capital flight cannot happen.

[Rick B]

This is a bit of a strange post. Tax rates are conscious political decisions made by the party in charge (bad decisions are justified with colloquial references to faux scientific and economic nonsense like "Hausers Law"). Federal tax rates combined with income distribution and corporate revenues would suggest that the system is designed to collect between 15-20% rather than 15 - 20% arising out of the postulates of economic theory.

I don't understand the point of the post, is it supposed to be a history lesson?

A 1-2% fluctuation in revenues for huge economies like the US is 100 - 200 billion dollars, or 3.5 - 7 million jobs. When a country is at 8.5% unemployment that is significant... For Canada its 10 - 20 billion dollars or 350 - 700 thousand jobs, and we also have high unemployment. Its a silly "Law" and its designed to minimize the role of tax revenues. I think the title is poor, and takes away from the post... maybe you were trying to be funny?

[Livio Di Matteo]

Don't quite understand the last comment. I've provided a historical perspective on the federal revenue to GDP ratio for Canada and the United States. It's postwar behavior for the United States has been referred to by Hauser's Law. I've provided stats for Canada to see if it might also characterize Canadian behavior. I 've been told I have a dry sense of humor but if any humor was intended in this case I'm sure it was entirely subliminal. Cheers!

[Rick B]

Sorry if I was unclear

I just sometimes get confused with the structure of some posts.

I appreciate the historical data, I'm just unclear as to why it was framed in terms of something like Hauser's law; when its not really a law, nor is it an argument, nor is it a theory. It just seemed like you were giving it credibility.

I just don't understand the context of the data presented that's all, nor do I get the intent of the post.