Larry Summers has an op-ed in the Financial Times arguing that federal expenditures are destined to increase as a share of GDP for a number of reasons, e.g.:
(1) The share of the population over 65 is on schedule to increase considerably, as is the dependency ratio, and 32% of the federal budget is now devoted to transfer programs for over-65s;
(2) when interest rates rise, debt service payments will increase;
(3) goods and services funded primarily by the public sector are becoming more expensive faster than goods and services paid for out-of-pocket;
(4) federal expenditures on infrastructure and pension liabilities might have to increase;
(5) as might spending on combating tax non-compliance.
(6) While technology might prove a boon, cash transfers and in-kind transfers represent a large and growing share of the budget. Another way of putting this is that as we reduce these transfers, lifetime net tax rates will generally increase.
My disagreements with Summers’ diagnosis are minor, but they lead me in a different policy direction. For example, I consider (3) both very important and not exogenous, i.e., there is reason to believe that the productivity- and value-enhancing business model innovation in the education and health sectors has been stymied by the outsized role of the public sector. On (4), there is a strong case for reorganizing responsibility for infrastructure investment, perhaps along the lines of the “Fix It First” strategy devised by Matthew Kahn and David Levinson.
On the subject of (3), Josh Barro has made worthwhile contributions, as in this Forbes post from late March:
In fact, the size of government ought to change as the economy changes. Some goods and services, like housing and food, ought to be bought mostly with private dollars. Others, like education and health care, should have a significant public expenditure component. As the public-heavy sectors grow relative to the private-heavy ones, government spending should grow, and vice-versa.
But while a growing health sector means that public expenditure should grow as a share of the economy, it also means that public expenditure on health care should shrink as a share of total expenditure on health care.
To be sure, there is room for disagreement on how large the public expenditure component in education and health care should be. More recently, Josh published a Bloomberg View post calling for universal coverage that is “basic” or “sufficient,” but decidedly not gold-plated.
There is a closely related subject that in my view doesn’t attract enough attention: while the public expenditures Larry Summers and Josh Barro reference may well have to increase as a share of GDP, I do think that there is a great deal of scope to reduce the size of the federal government’s various off-balance-sheet activities, e.g., credit guarantees. For a vivid example, please see Jason Delisle and Chris Papagianis on the FHA. These shadow expenditures are sufficiently large that curbing them — and, perhaps even more importantly, addressing overregulation — while allowing for a slight increase in on-balance-sheet activities might nevertheless yield a smaller government footprint. One serious potential concern is that when we focus exclusively on direct expenditures, we encourage the growth of government’s power via other less visible channels, a theme of Raghuram Rajan’s provocative Fault Lines.