As Jason Sorens notes, local government used to play a critical role in economic policymaking in the United States. While the federal government had a strictly limited role — to defend a national marketplace, to maintain the gold standard, and to distribute revenues from the national tariff — local governments engaged in a wide range of policies, from interventionist to laissez-faire:
Before World War I, not only was the federal government’s economic policy-making activity strictly limited to areas such as international trade, management of federal lands, trust-busting, and food and drug regulation, but state governments themselves were also internally decentralized. In 1913, local government own-source revenues (revenues raised autonomously by local governments, thus excluding grants) as a percentage of total state and local revenues (including federal grants to state and local governments) stood at a whopping 82%, according to my calculations based on historical Census Bureau data. If we assume that revenues track economic policy activity closely, this figure implies that four-fifths of all state and local economic policy activity occurred at the local level.
Today, of course, local governments are quite limited in their economic policy autonomy, with the most important remaining policy role left largely to local governments being K-12 education. Local revenue decentralization (the variable described in the last paragraph) was just 38% in 2008.
The key shift occurred during the Depression era:
In 1927, local revenue decentralization (LRD) still stood at 73%, nearly twice the level it is today. By 1932, near the nadir of the Depression, that had fallen to 68%. Under “that man in the White House,” the pace of centralization accelerated, as within two years LRD was at 59%, dropping further to 49% by 1940, not far from where it is today. To be fair, state politicians like Huey P. Long deserve some share of the blame or credit, depending on one’s perspective, for their attempts to set up state-level mini-New Deals. The federal role in centralization largely has to do with the expansion of federal grants to state and local governments, reducing the importance of autonomous revenue sources. The point is that within the space of about a dozen years, the U.S. political economy was transformed from one of extensive decentralization and robust local autonomy to one of federal dominance.
The remaining centralization episodes include the periods between 1940 and 1946 (-3% LRD), 1961-1969 (-6% LRD), 1971-1977 (-5% LRD), and 1991-1995 (-3% LRD). The latter two periods are probably associated with state supreme court decisions mandating state aid to local public schools, which in states like New Jersey and Vermont have been associated with noticeable jumps down in LRD. The escalation of the War on Drugs under Nixon, with its concomitant federal grants to local police departments, might also account for the 1970s trend. The 1960s centralization almost surely has to do with the expansion of the federal welfare state, which partly crowded out the pre-existing safety nets set up by local governments.
Sorens sees this centralization as an unfortunate development, as it undermines the ability of citizens to vote with their feet and, more interestingly, it reduces public interest in local affairs, thus allowing what is left of local government to fall prey to rent-seekers. This argument resonates in some respects with David Shleicher’s arguments concerning the role of national political parties undermines political competition at the local level, which Harvard Law professor Heather Gerken has described as follows:
He argues that there’s no reason to think that parties couldn’t fight about how to run local governments, which preside over such highly salient issues as local taxes, school funding, and crime prevention. The problem is that we don’t have enough media or campaign spending in local elections to make local party brands meaningful. Voters thus turn to the best heuristic they have—national party ID. Schleicher notes, for instance, that no matter what is happening locally, in most places votes in local elections roughly track votes in national elections. Schleicher bolsters his argument by identifying the instances where this isn’t true—where local parties manage to differentiate themselves from their national counterparts. Those examples tend to occur in places like New York City, where there is enough media coverage and campaign spending for mayoral candidates to create a distinctive local brand. Rudy Giuliani can escape the negative effects of the GOP label and win in heavily Democratic New York, but he would be stuck if he were running for mayor in New Haven. [Emphasis added]
These explanations aren’t mutually exclusive. The decline in local resource decentralization could, per Sorens, tend to reduce the amount of media coverage and campaign spending devoted to local elections because the stakes are seen as relatively low. And Schleicher’s mismatch — the use of national party heuristics in local contexts — could make matters worse by further reducing the public’s attention to what members of the local party elite are actually doing, e.g., the size of the concessions they are making to public workers. Consider, for example, the fact that in New York city, where media coverage and campaign spending are relatively high, average annual pension contributions have increased by approximately $7 billion over the last decade, yet the potential crowding-out of other valuable public services is never raised as an issue by potential mayoral candidates.
Sorens is brilliant, but his vision of a more libertarian U.S. does not comport entirely with my own. Though I’d favor more somewhat more local resource decentralization, I tend to think that reducing the volatility of state and local tax bases should be a high priority and that redistribution should generally occur at higher rather than lower levels of government, which is in tension with Sorens’s Tiebout-friendly vision of local government competition.
That is, I think that the federal government should tax mobile tax bases, like corporate income and nonwage individual income like capital gains, to the extent that these things should be taxed at all. Local governments, in contrast, should generally focus on taxes on the unimproved value of land, which offer a relatively stable revenue base and that create incentives to improve the desirability of living and working in a particular locale without creating disincentives to work and to earn. Now, I don’t think that the federal government should mandate this outcome. Rather, it shouldn’t stack the deck in favor of having state and local governments use certain taxes, which it does through the instrument of the state and local tax deduction. Eliminating the state and local tax deduction is essential to achieving a healthier, more constructive balance between the federal and state governments.
One thing Sorens doesn’t mention, because it is not directly salient to his point presumably, is that state and local government spending has doubled since the Korean War era. Much of this spending, as Sorens suggests, flows from mandates at the federal and state levels. That seems like a a good lever to use: reduce policy centralization through the use of grants and incentives.
To be sure, there are some grants and incentives that conservatives like very much, e.g., assistance to local police forces, which Sorens alludes to above. Chronic mismanagement has led to revenue starvation in some of our most dangerous, crime-ridden cities. At the same time, federal drug policies and and the spread of mandatory minimum laws at the state level have contributed to the larger incarceration crisis that has arguably exacerbated the crime problem. It’s hard to know where to begin unraveling this complex historical process that’s led to such terrible outcomes.