The government-spending effects of Progressive Era reform charters

There has been a boom of late in estimating the policy effects of historic local-government reforms. Those were often tied to forms of preferential voting. I want to share what I found when I did my own study of spending effects.

The takeaway is: higher aggregate spending under a council-manager charter that included single transferable vote (STV), compared to cities with non-STV manager charters, as well as cities whose ‘form’ of government didn’t change. I argue that this is to be expected, based on how the STV-manager charter combined the logics of “neighborhood representation” and “citywide focus” (both of which are common phrases in reform-practice circles).

The variety of reforms

Progressive Era municipal reforms came in three main flavors. One class strengthened directly elected mayors relative to local assemblies. It is not common to call this a “reform charter.”

Two more inaugurated the infamous ‘at large’ election. (On infamy, see Trebbi et al. 2008 or this new paper by Grumbach et al. 2023.) These also seem to have reduced the sizes of local assemblies.

One of them, the commission form of government, attempted to hold citywide elections to a series of functionally defined offices. Over time and to varying degrees, this morphed into a ‘numbered post’ electoral system. My former research assistant Andrew Rosenthal and I found nearly 100 such cases that came with an early form of ‘instant runoff.’ It is clear from historic advocacy materials that popular interests in preferential voting and more businesslike administration reinforced each other in helping commission government to spread.

The second at-large charter was designed, in part, to be compatible with some form of proportional representation (PR). It also aimed to piggyback on the joint popularity of preferential voting and businesslike administration. This was the council-manager form of government. It achieved compatibility with PR by junking the numbered posts.

Recent research

Two recent papers supply the impetus for this post. One from Carreri et al. (2023) finds limited effects on a range of inequality measures (pp. 13-14) from at-large-charter adoption, 1900-40. Another from Sahn (2023) disaggregates the two reform charters and points to greater capital outlays, 1900-34, from council-manger adoption. This is important; recall the “citywide focus” that council-manager is said to promote. (Also see Hankinson & Magazinnik 2023 on housing policy under at-large elections.)

There probably are other recent papers. I do not follow the historic urban political economy literature as closely as I might. I got to the topic by trying to say something about the effects of the reform I was studying (STV as embedded in a municipal reform charter).

Party competition in nonpartisan elections

My analysis used the data from this paper and focused on the period from 1930-60. This is important because, in 1932, the National Municipal League (NML) formalized the following policy: urge local reformers to form a “good government” party in any city where NML also was promoting council-manager. So, when we estimate the spending effects of these institutions, we may be picking up effects of the styles of party coordination they engendered.

It is worth noting that the new STV-manager charter in Portland (OR) once again has local reformers figuring out how to organize slates.

I argued that STV’s collision with a reform charter should lead to higher per-capita spending than non-STV charters and the absence of reform. The reason is that STV turned “good government” slates into “parties of neighborhoods.” The openness of an STV election made it unwise for reformers (now seeking to control government) to ignore (or harm) any given neighborhood. Evidence that politicians were thinking along these lines appears in the chapter. It includes first-choice vote shares by ward (when available), stories about campaign strategy, and stories about legislation (spot zoning, holding up slum clearance, etc). There also is a comparative literature on “localism” in STV elections (see, e.g., Carty 1981 on Ireland).

Here is the key table. “Plurality charter” really means “non-STV manager charter.” Unreformed city is the reference category. What this means is that the data do not record a change in form-of-government.

What we see above may combine what I’ve said about STV, plus the capital-outlays finding from Sahn.

The data in the other papers is better than what I used. Diff-in-diff also went into convulsion as I was writing my chapter, so I fell back on two-way fixed effects. I also did an RDD with pre-1930 data at one point, and that turned up no effects.

Awesome clams

Awesome clams as prepared on January 7, 2021.

Ingredients

  • 12-18 littleneck clams, cleaned of sand
  • 1/2 stick pepperoni or spicy chouriço
  • Medium onion
  • 2-3 cloves of garlic
  • Olive oil
  • Dry white wine or vermouth

Dice the sausage. Chop the onions finely. Slice the garlic. Heat the pan, and add oil.

Fry the sausage until it’s crispy. Add and sweat the onions. Add the garlic at some point. Experiment with before/after the onion sweating.

Deglaze with a splash of wine. Add more wine, and get it simmering.

Add the clams. Tuck them into the onion/sausage so that those aren’t on the bottom the whole time.

Cover the pan, and let it roll until all clams are opened.

Finished product, good with Portuguese rolls or crusty bread.

MPONP two years in retrospect

It has been two years since I finished/sent in More Parties or No Parties. Three issues with the theory have been on my mind. To recap, the theory says that “reform is common in periods of realignment” (or something along those lines). Then it says that a reform episode can be any of three types: insulating (comes from incumbent coalition), realigning (out-of-power folks peel off portion of incumbent coalition), and polarizing (opposing sides collude). The types were strictly internal to the polities whose institutions they were changing (cities for me, countries and some states for the lit review).

I do not have a thesis statement. I am making a list and riffing on it.

The first issue concerns coalition shift in general. I had cast this as resulting from activation of a coalition’s internal disagreement(s). Might not the addition of new players also destabilize a party system? (I think I acknowledged this possibility in a footnote.) Or might not the emergence of new issues do the same? I am thinking here of climate change especially, which is a ‘shock’ par excellence. It also has all sorts of implications for the issues that otherwise define the ‘basic space’ of a party system. What do we do when the ocean swallows the Outer Banks of North Carolina? Move the rich? Move everybody?

Another issue concerns the trigger(s) of a ‘realigning’ reform episode. Does the reform cause realignment, or does realignment cause the reform? The book argues that realignments cause reform, but a ‘realigning’ reform brings about (facilitates) realignment. So, which comes first? I suggested in ch. 8 that coalition shift at a higher level of government might produce demand for a realigning episode at some lower level. That would help to integrate other accounts of the reform adoptions I studied. For example, one good BA thesis argued that the Worcester reform charter was a business-community reaction to the rising power of the Irish — an insulating episode in my typology. I suggest it’s not this simple; the reform coalition included many people (Democrats) who helped elect a Mayor on the same day they ratified a charter meant to destroy his party. The key questions for me are about the magnitude and rapidity of change in advance of the reform. How big had the local Democratic Party gotten, and why? Maybe it comes down to an expanding set of salient issues, regardless of whether that’s at a higher level of government. I did find suggestive evidence with respect to unions. Maybe these are the same thing, at least sometimes.

Hopkins’ (2018) theory of nationalization is something I wish I’d brought in. Nationalization (as I integrate it into my mental model) means that higher-level cleavages are reshaping lower-level party systems. Or, in terms that are closer to those of Hopkins, national-level issues are more salient (for most people) to local politics than local-level issues. This leads (in my mind) to ‘oversized’ parties in some places, so that basic economic issues get debated within the oversized party. Witness non/bipartisan claims about corruption and how to pave a street.

Moving on.

I’m also not thrilled with having cast New York City (1936) as a ‘realigning’ episode. I am not saying this was wrong. The complication was the local separation-of-powers system. At the time of the reforms, Tammany controlled the Board of Aldermen, but LaGuardia controlled the mayoralty. And LaGuardia had won (and presumably controlled) a citywide vote majority.

So, maybe, the NYC reform was about creating congruence of control between the branches of government. (What the NYC reform package did to legislative power is fascinating, but let’s save it for another day.) Here is the article one might cite, then the key line: “Politicians became advocates of direct democracy, we argue, only where they were confident that voters were likely to agree with them and where current institutional arrangements blocked the median voter’s influence” (emphasis mine). Another way to put that: existing institutions (like malapportionment) were stopping statewide majorities from translating that strength into seat shares. That article is about different reforms, but the theory seems helpful. I also wonder if the identity/priorities of the ‘median voter’ changed in the period under our shared consideration (1898-1918). Otherwise, how would it have been possible to get around the incumbent coalition?