Economist Writing Every Day

Where would you expect Federalism to occur? In other words, where would expect a government to devolve authority to a lower government. Importantly, this is different from freedom vs authoritarianism. The lower government might choose to be more or less free. For example, right now in Florida there is a state-wide constitutional amendment on the ballot that would enshrine each individual’s right to hunt and fish. Ignoring the particulars of what that means, it’s clearly a step toward centralizing policy rather than decentralizing it. Central governments can be strong and protect citizens, or they can strip us of rights. Either way, being small players and far-removed, it’s difficult for us to affect the policy decisions.

That concern is philosophical, however. Maybe my opinion shouldn’t matter (one could easily argue). Even as a matter of prudence, one-size-fits all sets a standard, but the standard may not be a good fit for every locality and circumstance. There is a trade-off between ease of navigating a uniform policy across the land and customized policies that are particular to local priorities. Given that Americans can vote, is there a way for us to think about when a policy will be (should be?) centralized vs decentralized?

There is a great case study by Strumpf & Oberholzer-Gee* on the matter of alcohol policy after the end of national prohibition. The US has a dizzying array of liquor laws across the country and even across states. Some states have a central policy of dry or wet, while others devolve the authority to lower governments. How should we think about that policy? What determines the policy of central versus devolved authority?

One answer is variety. If people in a state are overwhelmingly either drinkers or teetotalers, then so goes the policy. But if there is a mix closer to 50-50, then states opt to avoid the political battle and devolve the authority.

But how do we know people’s preferences? We can examine popular election results or congressional voting records, but elections are infrequent and turnout is famously tricky due to strategic voting and arms races (why should I vote if the opponent will definitely win? why should I vote if my guy is a shoe-in?). If only there was some way of knowing people’s values prior to the policy coming up for vote…

Well, opinions on alcohol have historically been closely tied to religious affiliation. Since most people in the US were some sort of religion if not Christian, then knowing that someone was a Catholic or Lutheran was a pretty good clue that they would also be a drinker. Similarly, knowing if someone was Baptist, Presbyterian , or Methodist was a good clue that they didn’t drink. We have historical religious affiliation by county, which means we also have an indicator for preferences

This is exactly the link that Strumpf and Oberholzer-Gee exploit. They found that more Catholic states tended to have ‘wet’ policies at the state level and that more ‘Baptist’ states tended to be more ‘dry’ at the state level. Where there was a more even split according to alcohol-drinking religious affiliation, states chose to devolve authority to the counties. While causal inference wasn’t used, the theory and evidence is pretty strong. We can find confirming evidence at another level of analysis. In states with devolved authority, Catholic counties were wet and Baptist counties were dry. And more evenly mixed counties had semi-wet/dry policies. So it seems like the relationship holds.

There are also some nice results about the possibility of side-payments. You could imagine that a decisive local teetotaler might tolerate a wet county policy because he’s enticed by the prospect of enjoying the local tax revenue. The more that the local budget is funded locally rather than by the state, the more we should observe relatively ‘dry’ voters opting for wet policy locally. Similarly, the same decisive ‘dry’ voter loses the incentive to permit local liquor sales if a higher proportion of the county budget is received from the state. In such a case, the local budget is less dependent on alcohol sales and more apt to be dry. This pattern is observed too.

While alcohol prohibition and dry policies are not particularly of interest today, they are a revealing case study in how citizens of the past dealt with divisive issues. States received the power after national prohibition exactly because there was so much diversity of opinion. When there were loud and passionate voters on both sides with similar popularity, states decided to devolve the authority even further. This is what a functioning federalism looks like. If you look at some recent policies through this lens, then abortion rules being devolved to state control is a win for federalism. Increasing the cap on the State and local tax deduction helps make localities more responsive to the desires of their residents lest they migrate away. Under the last two years of the Biden administration, legal immigration has increased substantially. That greater diversity will push us in the direction of more federalism. It’s not always pretty, but these are steps in the direction of robust institutions.

---

*Strumpf, Koleman S., and Felix Oberholzer-Gee. 2000. “Local Liquor Control from 1934 to 1970.” In Public Choice Interpretations of American Economic History, edited by Jack Heckelman, John C. Moorhouse, and Robert M Whaples, 1–20. Boston: Kluwer. https://www.amazon.com/dp/0792377214?ref=ppx_yo2ov_dt_b_fed_asin_title.

*Strumpf, Koleman S., and Felix Oberholzer‐Gee. 2002. “Endogenous Policy Decentralization: Testing the Central Tenet of Economic Federalism.” Journal of Political Economy 110 (1): 1–36. https://doi.org/10.1086/324393.