Larry Summers Conflates Economics and Politics on Shrinking Government

Former Treasury Secretary Larry Summers claimed at a Wednesday lunch that the “Republican vow to significantly reduce the size of government is a foolish pipe dream” due to “structural economic realities.”

What are these realities, according to Summers?

  1. An aging population will mean upward pressure on entitlement spending on unchanged policy.
  2. The rise in inequality requires government spending to “ameliorate” the consequences.
  3. Prices tend to rise relatively quickly in service sectors such as education and healthcare, necessitating more government spending.
  4. Rising national security threats and increased military spending by geopolitical foes will necessitate more U.S. military spending too.

Where to start?

Summers is right that, on unchanged policies, government spending would balloon due to aging.

The Congressional Budget Office projects spending on Social Security would rise from 4.9 to 6.3 percent over the next 30 years, whilst Medicare spending would nearly double from 3.1 percent of GDP to 6.1 percent. The impact of all that extra spending, even as non-Social Security and healthcare spending is projected to fall from 8.9 percent of GDP to 7.6 percent, is a growing budget deficit and accumulated debt. This would raise net debt interest payments further, such that by 2047 the U.S. budget deficit stood at (a completely unsustainable) 9.8 percent of GDP.

These long-term projections come with all the usual caveats. They use assumptions about the likely path of productivity growth in the economy, population growth, and the extent of labor force participation. Nevertheless, this analysis does highlight the scale of the contingent liabilities embedded in current policy.

Summers is also right about movements in relative prices, at least if historic trends continue. Labor intensive service industries, particularly where government involvement and support is high, seem to have seen price explosions relative to general price indices. Absent a future productivity take-off in medicine and education, it is certainly plausible that government will be under pressure to spend even more in these areas to maintain services.

The main problem that I have with Larry’s argument is not his description of these economic phenomena. It’s that on the implications of these facts and his further points about inequality and defense spending, he conflates economics with politics. He is guilty of a category error, combining elements of positive economics with normative judgments, especially when it comes to necessity to tackle inequality.

If you assume (as others such as Will Wilkinson have argued) that the public will not accept a smaller government because they do not want it, that entitlements cannot be touched and that there is an income elasticity for government spending > 1 (Wagner’s Law), then the trends outlined above are likely to grow government. But this need not be the case from an economic perspective. There are plenty of other policy options available. In other words, Larry’s political judgments help drive what he describes as “economic reality”.

This reminds me of the Brexit debate, when multiple economic agencies, including the UK government’s own Treasury, proclaimed there could only be economic losses from the UK leaving the EU. Looking at their assumptions, one saw these results arose largely by construction. Assuming Britain would sign no new free trade deals and change no regulations, but still suffer reduced trade within Europe, then Brexit inevitably would make Britain poorer.

So what does Larry miss?

First of all, there’s a whole host of areas where the federal government could reduce spending outside of healthcare and Social Security spending. Though quantitatively some of these areas might be small, and others would be politically difficult, qualitatively they would reduce the scope of government substantially. Chris Edwards has outlined a potential $457.8 billion in savings in the recent Cato Handbook for Policymakers.

Second, though aging does put upward pressure on spending, the above figures show that the current entitlement framework is unsustainable without mammoth increases in the tax burden, which would reduce the potential growth rate of the economy. Entitlement reform in itself then is an exercise in “downsizing government” relative to current policy, and only reform here can maintain or reduce the size of government overall. Larry might be right that entitlement reform is politically difficult, but that is not an economic argument.

Third, and finally, though there are other economic explanations for the rising relative prices of education and health services, it seems likely that at least part of the story is precisely the role of administrative bloat and lack of competition brought about by government involvement. If so, then Larry’s argument gets things the wrong way round – the necessity of more government spending will be (in part) a consequence of government interference in the first place.

As my colleague Dan Mitchell has acknowledged, downsizing government is no doubt a difficult endeavor. That’s one reason why Cato has a whole stream of work dedicated to it. But Summers should not pretend that political choices are the same as economic realities.