Voting With Your Feet
POLS 4641: The Science of Cities
Warmup
Wait. What?
The US Census of Governments defines a government as “an organized entity which, in addition to having governmental character, has sufficient discretion in the management of its own affairs to distinguish it as separate from the administrative structure of any other governmental unit.”
- Basically, any entity (1) with defined geographic boundaries, (2) that can levy taxes, and (3) has its own elected officials.
This includes, approximately:
20,000 cities
16,000 townships
38,000 special districts
13,000 school districts
3,000 counties
50 states
1 federal government
Today’s Agenda
Is this too many governments? Too few?
- What are the consequences of dividing political authority among so many independent entities?
We’ll explore this question through a series of simulations, each illustrating a few key theoretical ideas.
Simulation 1
Simulation 1: Setup
Imagine that you are a person trying to decide where to live.
You only care about three things:
- Playing pickleball.
- Reading books at your local public library.
- Not paying too much in taxes.
Everyone take a sheet of paper, which describes exactly how much you care about each of these things.
Example Set of Values:
| People per Pickleball Court | Value |
|---|---|
| 10 | $100 |
| 5 | $190 |
| 2 | $220 |
| 1 | $220 |
| Books in Library | Value |
|---|---|
| 0 | $0 |
| 1 | $30 |
| 2 | $50 |
| 3 | $65 |
Simulation 1: Towns
In order to get pickleball and library books, you need to organize yourselves into towns.
In each town, choose one person to be mayor. Ideally someone trustworthy. Someone who doesn’t mind doing a little arithmetic.
The mayor is responsible for choosing the town’s policy bundle.
- The “lump sum” tax rate (i.e. each resident must pay $X in taxes).
- How many pickleball courts the town will maintain.
- How many books will be in the town’s public library.
The mayor will receive a sheet of paper describing the costs for each level of public goods.
- You must raise enough in taxes to cover your costs!
Simulation 1: Sorting
You may, at any point during the simulation:
Move to a different town (“migration”).
Create a brand new town, declaring yourself mayor (“secession”).
Dissolve a town where you are mayor and join another town (“amalgamation”).
Mayors may freely adjust their town’s policy bundle in response to new migration, secession, or amalgamation.
You have 10 minutes to make yourselves as happy as possible.
Go!
Simulation 1: Debrief
What happened?
This simulation captures a few key ideas from (Tiebout 1956). (Pronounced “TEE-bow”).
- Competition between local governments is the public-sector analogue to the private-sector market.
- An efficient outcome can be reached through decentralized, individual decisions. No politics necessary! 😅
- In a diverse polity, people with different preferences for public goods can sort themselves into jurisdictions best matching their preferences.
- Because public goods exhibit economies of scale (Blom-Hansen, Houlberg, and Serritzlew 2014), local governments have an incentive to compete to attact new residents, lowering average costs.
- Like private firms, local governments are pressured to keep costs (taxes) low in order to successfully compete for residents.
- Competition between local governments is the public-sector analogue to the private-sector market.
What important things are left out of this model?
Simulation 2
Simulation 2: Adding a Few Wrinkles
Now we’re going to run the same simulation again, but we’ll add three new rules:
- Everyone now has different levels of income. Grab a card to figure out how much money you have.
- Mayors: instead of “lump sum” taxes, you must now set tax rates in percentage terms. What percent of each resident’s income are you taxing?
- It costs $50 to move from one town to another (or to secede/amalgamate). Just deduct it from your money total whenever you move.
That’s it! I’ll give you another 10 minutes to make yourselves as happy as possible.
Go!
Simulation 2: Debrief
What happened?
Perhaps things played out a bit differently when we added some friction to the idealized model:
- When income is heterogeneous, taxes are inherently redistributive.
- This pressures people to sort not just on preferences, but on income as well.
- Moving costs limit the ability of poor households to integrate with rich households.
- In practice, such costs include more than just the physical costs of moving (Hamilton 1975; Derenoncourt 2022).
City Limits
Peterson (1981) argues that these forces place fundamental constraints on the sorts of public spending that is feasible for local governments.
He distinguishes between three types of public spending:
- Developmental (spending designed to attract private investment)
- Redistributive (transfers from rich to poor)
- Allocative (everything else; basic services)
Capital mobility makes redistribution tough to finance through local tax revenue.
Simulation 3
Simulation 3: Setup
This one will be quick.
Reconvene with the members of your town. You have one last decision to make.
I am a multinational corporation, and I am considering investing in your town. Congratulations!
I have $10,000 ready to invest right now. (Think of all the pickleball courts that could buy!)
You just have to decide what percent of that investment you’d like to tax.
Take 5 minutes to deliberate, then secretly submit your best offer. I will split my investment evenly among all the towns that offer me the lowest tax rate.
Go!
Simulation 3: Debrief
What happened?
I don’t know what happened. I’m writing this bullet point in the distant past!
But…here’s what theory says should happen?
If every town can successfully coordinate to offer a tax rate of, say, 10%, then every town gets \(\frac{\$1,000}{n}\). Hooray!
But if any town breaks that agreement, offering a tax rate of 9%, they get the entire investment ($900 in tax revenue)!
This is a classic Prisoners’ Dilemma.
| Tax at 10% | Tax at 6% | |
| Tax at 10% | $500, $500 | $0, $600 |
| Tax at 6% | $600, $0 | $300, $300 |
- And the logic doesn’t stop there. Why not offer 5%? 3%? 0.0001%? This is a Race To The Bottom.
Key Takeaways
Is 90,000 too many governments? Well, it depends what you want to achieve…
The Tiebout (1956) model (Simulation 1) suggests that some types of public spending are best handled by competing local governments:
Lots of preference diversity
Limited economies of scale
No “spillovers” across geographic boundaries
Residential mobility limits the types of public spending that is feasible for local governments, particularly redistribution (Simulation 2).
Tax competition for mobile capital can yield a “Race to the Bottom” (Simulation 3).
Scholars disagree on how good a model this is of local tax incentives (Jensen, Malesky, and Walsh 2015).
But it’s a very good model of the competition for sports stadiums!
More on that in the Deep Dives.