My Economics 21 colleagues have highlighted an often overlooked fact about the growth of public spending as a share of U.S. GDP in a brief primer on state and local public sector compensation:
Measured from the end of the Korean War (1953) to the year before the financial crisis (2007), federal spending had actually declined as a share of GDP, from 20.4% to 19.6%. Since the 1960s, federal spending has exhibited no persistent upward or downward trend. Periods of rising federal expenditure have been followed by periods of declining outlays, measured relative to GDP. The federal government has grown substantially over the entire period, of course, but this growth has been roughly in line with that of the overall U.S. economy. By contrast, over the past 60 years, state and local government expenditures have doubled as a share of the economy, from 7.7% of GDP in 1950 to 15.5% in 2009. Since 1950, state and local spending has grown at an 8.1% annual rate, fast enough to double the size of state and local government every 8 or 9 years.
Perhaps this growth might not be so bad, depending on where those incremental expenditures are allocated. If the much higher spending were directed towards improved infrastructure, perhaps it could be considered money well spent. But that’s not the case. Data from the BEA shows that Gross state and local investment – spending on things like roads, hospitals, prisons, highways, ports, and transit systems – has virtually flat-lined as a share of GDP since 1950 while the rest of the state and local spending has increased by 130% since then.
While this won’t come as a shock to many of you, the increasing importance of state and local spending raises interesting questions:
(1) As the article goes on to observe, spending on the state and local workforce is an important component of the spending increase. Yet it’s not clear that steady decreases in the student-teacher ratio in U.S. public schools has been the most effective use of public resources, or that state governments have achieved an ideal, flawlessly efficiency-enhancing mix of compensation schemes. (This is part of what I find so interesting and peculiar about critics of compensation reform: do they really believe that there is no room for improvement? Critics of PPACA almost never claim that there is no room for improvement in the U.S. health system, which is why increasing energy is being devoted to ideas about how to “replace” the new health law.)
(2) Why isn’t Tiebout choice restraining state and local spending growth more effectively?
(a) Decreases in marginal tax rates at the federal level have made state income taxes somewhat more salient, as the state and local tax deduction is worth considerably less to high earners than it was in the 1960s and 1970s. This has helped encourage at least some restraint in terms of tax levels.
(b) But Medicaid’s federal-state nature has created misaligned incentives that have encouraged spending growth across all states, and the same is true of other smaller and lesser-known federal-state programs. When competitive federalism is replaced by cooperative federalism, it is almost inevitable that we’d see more convergence across state governments in terms of how spending programs are structured. That is, if states seek matching grants from the federal government according to federal funding formula, they will have to harmonize their approaches on at least some dimensions.
My sense is that (b) is a stronger force than (a). Moreover, tax restraint will not automatically deliver spending restraint if state officials are convinced that they can extract more money from the federal government or engage in shadow borrowing.
(3) As the country has grown more affluent and unequal, and as the demographic composition of the population has changed — more single-parent families, more single-occupancy households, etc. — demand for public services has increased. It’s not clear to me that this demand had to be met by public providers. Yet there’s no denying the increased demand. As we turned away from the forced institutionalization of the mentally ill (and those deemed mentally ill), a need emerged for more expensive and humane outpatient facilities, a homelessness problem emerged that had to be managed through the deployment of public resources, etc.
The advent of new narcotics like crack cocaine helped fuel a crime explosion, prompted a concomitant increase in incarceration levels. And even as the crack epidemic waned, one could argue that rising affluence influenced the demand for what we might call a higher level of public order. When murder rates decline, you don’t hear anyone say, “Well, in that case let’s shrink the size of the police force.” Some, myself included, argue that we should continue increasing the per capita number of police to drive down other violent crimes and further improve the quality of life. This is public spending.
(4) A big part of the story, I suspect, is that the relative attractiveness of state-level public sector employment has created a large and very loss-averse population of engaged voters. Moreover, the expansion of social scientific disciplines has been tightly linked to the expansion of the state, a subject to which I’ll return.