semi-automatic waffle

Yes. A useful way to approach this is to treat suffering and injustice not merely as failures of individual morality, but as properties that can emerge from incentive structures, institutions, and feedback loops.

The central question becomes:

How do we design institutions so that people can pursue their interests without systematically imposing large costs, risks, or losses of agency on others?

There isn't one institutional design that solves this. But there are several fairly general principles.

1. Align private incentives with social consequences

A recurring problem is that an actor receives the benefit while someone else bears the cost.

Examples:

A company profits while pollution is borne by the public.
A manager improves quarterly metrics while employees absorb burnout.
A platform maximizes engagement while users absorb attention and mental-health costs.
A government agency expands its budget by expanding the problem it is supposed to solve.
An individual gains status by behaving competitively even when cooperation would make everyone better off.

The remedy is not simply "make people more moral." It is to change the payoff structure.

That can involve liability, Pigouvian taxes, regulation, insurance, compensation, reputation systems, collective bargaining, transparency, or redesigning performance metrics.

A useful design heuristic is:

Whenever possible, make the person who creates a cost participate in bearing that cost—and make the person who creates genuine social value benefit from doing so.

But this has to be done carefully. Crude incentive schemes often create gaming and unintended consequences.

2. Measure what actually matters

Organizations frequently optimize what they can easily measure rather than what they actually care about.

A hospital can optimize throughput and inadvertently make care less humane.
A university can optimize publications and produce perverse research incentives.
A company can optimize quarterly earnings and sacrifice resilience.
A school can optimize test scores and undermine curiosity.

This produces Goodhart's law: once a measure becomes a target, it tends to stop being a good measure.

So institutions need plural objectives, including measures that are harder to quantify:

long-term consequences
quality
wellbeing
resilience
distribution of benefits and harms
autonomy
trust
environmental effects
effects on people who lack bargaining power

And some things should be treated as constraints rather than variables to optimize. For example, an organization shouldn't be permitted to compensate for severe violations of someone's rights by producing more benefits elsewhere.

3. Give affected people actual power

One of the deepest sources of institutional injustice is asymmetry of power.

If A makes decisions that affect B, while B has little ability to influence A, A has an incentive to externalize costs onto B.

This occurs inside corporations, bureaucracies, families, housing markets, international supply chains, and political systems.

One response is to introduce voice and countervailing power:

employee representation
unions
consumer protections
independent regulators
shareholder democracy
community consultation
ombuds institutions
accessible courts
freedom of association
whistleblower protection
participatory governance

The important distinction is between being consulted and having meaningful bargaining power.

4. Distribute power rather than merely redistributing money

Inequality is often discussed primarily in terms of income. But power inequality can reproduce itself even after income redistribution.

Someone who controls:

capital,
information,
employment,
housing,
infrastructure,
political access,
organizational decisions

may have enormous influence over people who technically have similar incomes.

Consequently, ameliorating injustice often means changing ownership and decision-making structures, not merely transferring resources after the fact.

This is one reason experiments with cooperatives, employee ownership, mutual organizations, public-interest institutions, and various forms of democratic governance are interesting: they change who has a stake in decisions.

5. Build institutions around human limitations

People aren't consistently rational, altruistic, patient, or self-controlled.

But neither are they simply selfish.

Humans are extraordinarily sensitive to:

status
reciprocity
fairness
belonging
loss
social comparison
authority
group identity
immediate rewards

A good institution therefore shouldn't require everyone to be exceptionally virtuous.

It should be robust to ordinary human psychology.

For example, instead of saying:

"Managers should never exploit their employees."

create conditions in which exploitation is difficult:

transparent pay + worker representation + labor mobility + enforceable employment rights + independent complaint mechanisms.

This is an important philosophical shift:

Don't design society for saints. Design it so that ordinary people can behave reasonably without extraordinary moral effort.

6. Reduce unnecessary zero-sum competition

Some competition is productive. But institutions can accidentally turn intrinsically cooperative activities into status competitions.

If 100 people need meaningful work but only 20 prestigious jobs exist, the system can make 80 people feel like failures even though society still needs their contributions.

Similarly, if social status is strongly tied to wealth, people may spend enormous resources competing for relative position rather than producing additional wellbeing.

This suggests designing institutions that provide multiple sources of dignity:

meaningful work
community participation
craftsmanship
caregiving
education
artistic contribution
scientific contribution
public service

The goal isn't to eliminate competition. It is to prevent every important human good from becoming a tournament with winners and losers.

7. Make exit possible

Voice is important, but so is exit.

An institution becomes dangerous when people cannot realistically leave it.

Examples include:

monopolistic employers
abusive relationships
authoritarian organizations
exploitative landlords
highly concentrated digital platforms
bureaucracies with no appeal mechanism

So a healthy institutional ecosystem often needs redundancy:

voice + exit + competition + due process.

If you cannot improve an institution from inside it and cannot leave it, you're effectively trapped.

8. Put hard limits around power

There are some domains where incentives should not be allowed to determine everything.

Markets are excellent at coordinating many activities, but societies generally establish boundaries around things such as bodily integrity, due process, political rights, and basic human dignity.

Likewise, democratic majorities need constitutional constraints because majority power can itself become oppressive.

So institutional design has a recurring structure:

Give people freedom to pursue their interests, but establish boundaries preventing them from using that freedom to destroy the freedom of others.

9. Create feedback mechanisms

Institutions become pathological when they cannot learn.

A healthy organization should be able to discover:

"Our policy produced an unintended harm."

and then actually change it.

That requires:

independent auditing
transparent information
whistleblowing
experimentation
appeals
error reporting
post-mortems
independent journalism
scientific evaluation
sunset clauses
periodic institutional review

This may be one of the most underappreciated principles.

A good institution isn't one that never makes mistakes. It's one that can detect and correct them without requiring catastrophe first.

A deeper principle: minimize domination

These ideas can be unified around a broader concept than equality.

Imagine three societies:

A: Everyone receives roughly the same resources, but a small group can arbitrarily control everyone else.

B: Resources are unequal, but everyone has strong rights, substantial bargaining power, and realistic opportunities to leave.

C: Resources are relatively equal, power is decentralized, and people have meaningful control over decisions affecting their lives.

Income statistics alone don't tell us which society is more humane.

A deeper criterion is:

How much arbitrary power does one person or institution have over another person's life?

This connects ideas from republican political philosophy, democratic theory, labor economics, institutional economics, organizational psychology, and human-rights theory.

There is also a danger in trying to "design society"

This is crucial.

The people designing institutions are themselves embedded in incentive structures. They have incomplete information, biases, interests, and limited knowledge.

Consequently, the ambition shouldn't be:

"Find the perfect institutional arrangement."

It should be:

"Build institutions capable of discovering when they're producing harm and correcting themselves."

That leads toward institutional pluralism and experimentation rather than one grand blueprint.

Different organizations can try different governance arrangements. Successful arrangements can spread; unsuccessful ones can be abandoned. People should have meaningful ability to exit. Power should be distributed sufficiently that no failed experiment can impose itself universally.

In other words, perhaps the deepest institutional design principle is evolutionary rather than utopian.

A compact framework

When examining any organization or social system, I'd ask:

Question	What it reveals
Who benefits?	Incentives
Who bears the costs?	Externalities
Who makes the decisions?	Power
Who has information?	Epistemic asymmetry
Who can leave?	Dependence
Who can object?	Accountability
What gets measured?	Optimization pressure
What gets ignored?	Hidden harms
What happens when someone behaves badly?	Enforcement
What happens when the institution itself is wrong?	Correctability
Can the powerful change the rules?	Institutional capture
Can affected people change the rules?	Democratic legitimacy

That framework can be applied at almost any scale—from a family or workplace to a corporation, university, city, or state.

And there is an interesting philosophical conclusion: justice may be less about creating perfectly equal outcomes than about constructing systems in which people have security, agency, voice, dignity, and protection from arbitrary power—and in which harmful arrangements can actually be changed.