In a recent post at EconLog, I discussed several phenomena that limit the effectiveness of electoral feedback. One of these is bundling. Each candidate in an election will have power to make policies that impact a wide variety of issues. Yet each citizen has just a single vote. So it’s not clear whether a vote expresses preferences about health policy, foreign policy, or any number of other policy arenas. Moreover, a voter might vote for a candidate whose foreign policy they dislike if they like the candidate’s health policy enough. This bundling means that electoral feedback is at best a very noisy signal about voters’ approval or disapproval of particular policies or actions.
Arnold Kling favorably quoted my post at his Substack, and made a very good point about how this bundling issue impacts not just electoral feedback, but also the feedback associated with interjurisdictional competition:
This is an excellent point! Bryan Caplan addressed similar issues in his recent plenary address at the Public Choice Society meetings, titled “Tiebout Was Wrong, but Why?” While Kling focuses on problems of bundling, Caplan largely focuses on the proposition that “non-profit competition is far inferior to for-profit competition. The ‘competition’ between local governments is like an academic test that doesn’t count for your grade.”
Kling addresses issues related to bundling and what that means for the options that citizen-consumers have. Whereas Caplan largely addresses issues related to the fact that political decision-makers are not residual claimants and therefore face very different incentives than entrepreneurs in a market do.
It may be possible to address both of these issues, though doing so is somewhat radical. To address Kling’s concern, we could consider unbundling various aspects of governance and allowing for non-territorial entry and exit by citizen-consumers. Trent J. MacDonald proposes this and discusses the idea in detail in his excellent book The Political Economy of Non-Territorial Exit: Cryptosecession, which I reviewed for the Review of Austrian Economics.
As I explained in my review:
So MacDonald specifically discusses unbundling and non-territorial exit as a way to address real-world limitations to Tiebout competition.
What about Caplan’s concerns about motivations, incentives, and residual claimants? Well, one thing that could be done is turning more local government services over to for-profit entrepreneurs operating in a competitive market. In the limit, this might take the form of market anarchism or anarcho-capitalism, in which even law and security are provided on the market!
There’s plenty of room for research on the sources of “quasimarket failure.” Perhaps one way to alleviate these failures is to take the “quasi” out of “quasimarket”!
READER COMMENTS
Jon Murphy
May 26 2022 at 10:03am
Great post.
I’m actually a huge believer in Tiebout competition. It recently played a role in my choosing where to take a job (I liked the bundle of government services and taxes better in NC than NY) and I am working on a paper where I use Tiebout to explain the ineffectiveness of vaccine mandates. But the points you, Kling, and Caplan all make are very valid. I think it’s a very rich area for research.
robc
May 26 2022 at 10:35am
But wouldn’t it have been even better to order a la carte? Choose the best of NC and the best of NY.
Jon Murphy
May 26 2022 at 10:37am
Absolutely.
Christophe Biocca
May 26 2022 at 10:42am
Private jurisdictions that are aware of the bundling problem can also just allow for competition with their own services without needing to go full anarcho-capitalist.
That is, they can keep their territorial monopoly on a handful of narrow slices while allowing subgroups to opt out of their default provision road maintenance, schools or whatever else. Próspera takes this to an extreme, allowing people to pick regulatory codes from existing OECD countries, and pick their own insurer to act as first-line regulator. So a hospital running under Swedish regulations can buy cleaning services from a company operating under US rules, and so on.
This makes sense if you’re a residual claimant, as making the jurisdiction more attractive is much more valuable that whatever surplus you could extract from a monopoly on curbside garbage pickup. In that sense the bundling problem is downstream from the incentive problem.
robc
May 26 2022 at 11:39am
I think school choice is a classic example that is actually being done.
My daughter attends and out-of-district charter school. Covid killed off any waiting list, making it easy to get in last year. And now that she is in, she has a spot going forward. In 3 years, we will probably be switching to a charter school within our district.
But the biggest deal is not being tied to a specific neighborhood school.
Philo
May 26 2022 at 6:03pm
The bundle you face includes not only local government services but also the local people, with their cultural characteristics (and possible family relatedness to you), and the local geography and climate. Even if the cost of moving were negligible, the chance of finding a bundle of all this stuff that was just what you wanted would be small.
robc
May 26 2022 at 8:10pm
If you could shop for local government services, choosing based on geography would be much easier.
Matthias
May 27 2022 at 11:02pm
Without going too radical, one obvious method to make Tiebout competition stronger is to push more political decision making more local, to smaller units.
There are more towns than countries. And it’s easier to move between towns, too.
Subsidiarity is a great principle.
robc
May 28 2022 at 10:12am
I don’t think that saying that most government decision making should be made at an organization about Dunbar’s Number size is radical. It is based on some reasonable social science.
The fact that most current governments are larger than 150 people makes it seem radical.
Ian Fillmore
May 31 2022 at 2:19pm
It seems to me that the effectiveness of Tiebout competition depends directly on the richness of the “product space.” Ideally, every conceivable bundle would be offered, and everyone could pick the bundle he or she liked. In reality, only a relatively small set of those bundles get offered. I think a fruitful line of research would be understanding why, in equilibrium, only a small set of bundles get offered.
In the electoral context, the median voter theorem nicely illustrates why only a few bundles would be offered (and those should be pretty similar to each other). The inter jurisdictional competition case is much more interesting, I think. Something to chew on…
Comments are closed.