the market for political certainty

what if the behavior that helps a politician succeed isn’t always the behavior that helps society succeed?

it sounds almost too obvious to be interesting.

yes, politicians have incentives. yes, elections reward some things more than others.

but once you start taking the question seriously, it becomes difficult to stop thinking about it.

because many of the things we complain about in politics might not simply be the result of individual character.

they might be predictable responses to the system we’ve designed.

and if that’s true, demanding better politicians might be necessary.

but insufficient.

1the problem with asking people to be better

imagine momo and lolo are discussing a politician who has just announced a major infrastructure project.

momo likes the ambition. the city desperately needs investment, and she’s tired of governments announcing consultations instead of actually building things.

lolo is more skeptical. the project looks expensive, its benefits are uncertain, and she’s worried that politicians will celebrate the announcement long before anybody can evaluate the result.

both have a point.

the announcement is visible today.

the cost will be spread over years.

the benefits might not become clear for a decade.

and by then, the politician who announced it may no longer be in office.

so whose incentives are we actually looking at?

the politician’s?

the voters'?

the government’s?

the people who will use the infrastructure?

the taxpayers who will finance it?

the people who haven’t been born yet?

these are not (exclusively) the same people.

which means their interests don’t have to align.

2politics as a principal–agent problem

in economics, a principal–agent problem arises when one party delegates a task to another, but cannot perfectly observe or control the agent’s behavior.

think of asking a friend to water your plants while you’re away for a month.

you want the plants to be healthy, long after you’re back.

your friend wants that too.

but they also have their own life, a busy week, and one moment by which they’ll be judged: the day you walk back through the door and look at your plants.

you can’t see what happens in between. you only see the result, on one particular day.

so it’s tempting to forget them for three weeks and drown them the evening before you return.

even when everyone is well-intentioned, incentives can diverge.

democratic politics has a related problem.

citizens delegate authority to elected representatives.

but citizens cannot continuously monitor every decision, understand every technical trade-off, or directly observe the long-term effects of every policy.

and politicians are evaluated through elections that take place at specific moments.

we can take this simple objective function for a hypothetical politician:

where:

• V is electoral success;
• R is reputation, influence, or political standing;
• W is the politician’s valuation of public welfare;
• K is the personal or political cost of pursuing the policy.

this isn’t an accusation that politicians only care about themselves.

the model explicitly allows politicians to care deeply about public welfare.

it’s simply a reminder that they operate under several incentives at once.

and those incentives can conflict.

the crucial phrase is incomplete, delayed results.

political accountability is real.

but it is not a perfect measurement system.

3the tyranny of the visible

consider two possible policies.

policy a creates a visible benefit before the next election, but generates significant costs later.

policy b imposes an immediate cost, but creates larger benefits ten years from now.

it’s entirely possible that policy b is better for society.

and policy a is better for the politician’s short-term prospects.

this is not inevitable. voters sometimes reward difficult reforms, and institutions can support long-term decisions.

but the tension exists.

let’s make it concrete with a toy model.

a policy costs 20 units of welfare today and generates 40 units next period.

if society values next-period welfare almost as highly as present welfare, using a discount factor of 0.95:

the policy creates positive net present welfare under these assumptions.

now imagine that, for electoral reasons, the politician receives only 20% of the political credit for benefits that arrive after the next election:

same project.

same physical consequences.

different incentives.

the numbers are invented, and political credit isn’t literally welfare. to me, that’s the point: we’re comparing two distinct objective functions.

the model doesn’t prove politicians avoid long-term reforms.

it shows why the timing of benefits and costs matter.

the upper line uses a 0.95 discount factor; the lower uses 0.55. these are deliberately stylized preferences, not measured societal or political discount rates.

and the obvious question becomes:

how do we design institutions that make it easier to do something beneficial even when its rewards are delayed?

4why information makes the problem harder

now add another complication.

voters rarely observe policy quality directly.

they observe announcements, prices, headlines, public services, economic indicators, and their own experiences.

all of which matter.

but outcomes are influenced by many things beyond one politician’s control.

an economy can perform well despite a bad policy.

or badly despite a good one.

a government can benefit from favorable global conditions.

or suffer from a crisis it didn’t create.

this makes accountability noisy.

and when performance is difficult to observe, communication becomes especially important.

sometimes for good reasons.

a government has to explain its decisions.

but communication can also become a substitute for evidence of results.

it may be easier to announce a reform than to implement it.

easier to claim an achievement than to measure its long-term consequences.

easier to blame an opponent than to explain a complicated failure.

none of this requires believing that all politicians are cynical.

it requires only recognizing that attention and attribution are scarce resources.

5the public choice perspective

there’s a tradition in economics called public choice theory.

one of its central contributions is to apply the analysis of incentives to political actors rather than treating government as a single benevolent decision-maker.

politicians respond to incentives.

bureaucracies respond to incentives.

interest groups respond to incentives.

voters respond to incentives.

incentives matter.

and different institutions create different patterns of behavior.

this can be a useful corrective to a common mistake.

we often compare an imperfect market with an idealized government.

or an imperfect government with an idealized market.

neither comparison tells us much.

the relevant question is how actual institutions perform, given actual information problems, incentives, constraints, and human behavior.

i think that’s an important intellectual habit.

especially for someone who believes both in the potential of markets and in the need for a capable state.

markets can fail.

governments can fail.

and simply identifying one failure doesn’t automatically establish that the other institution will solve it.

6but public choice can become a lazy form of cynicism

there’s a trap here.

once you start explaining politics through incentives, it’s tempting to assume that every action is secretly self-serving.

a politician supports a reform? they must be chasing votes.

a civil servant resists a proposal? they must be protecting their bureaucracy.

a voter advocates redistribution? they must want someone else’s money.

that kind of reasoning quickly becomes unfalsifiable.

and uninteresting.

people have principles.

they care about other people.

they sometimes act against their narrow material interests.

institutions also contain professional norms, legal obligations, and reputational mechanisms that encourage responsible behavior.

the fact that an action is compatible with self-interest doesn’t prove self-interest caused it.

a good model should make us more curious about motivations.

not less.

7what better incentives might look like

if the problem is partly institutional, the response has to be partly institutional too.

we might ask how governments can make long-term policy effects more visible.

how independent evaluation can improve accountability.

how transparent budgets can reveal costs that would otherwise be deferred.

how administrative capacity can turn political promises into measurable outcomes.

how voters and journalists can distinguish an announcement from implementation.

and how democratic institutions can make it possible to correct a failed policy without treating every correction as a confession of moral bankruptcy.

these aren’t magical solutions.

independent institutions can make mistakes.

metrics can be gamed.

technocrats can be unaccountable.

long-term planning can become an excuse to ignore democratic preferences.

but at least these questions focus on the architecture of the system, not only the virtue of the individuals inside it.

8the problem with rewarding the wrong thing

momo and lolo are still arguing about the infrastructure project.

eventually they agree on something.

the question shouldn’t be whether the announcement sounds ambitious or whether the politician belongs to their preferred party.

the question should be what the project is supposed to achieve, what it will cost, how success will be measured, and when they’ll know whether it worked.

they may still disagree about whether it’s worth doing.

that’s fine.

but they’ve moved from judging the performance of politics to evaluating the substance of policy.

and perhaps that’s the distinction we need more often.

because if political success is measured mainly by the ability to win arguments, dominate attention, and survive the next election, we shouldn’t be surprised when political actors become exceptionally good at those things.

the more difficult challenge is creating incentives to be good at something else.

governing.

opening another question.

even if a politician has excellent policies, why do some political movements struggle to make people believe in a better future?