[I am in a group chat with the author and asked permission to repost here. He's not on EA forum. But I think the post has ramifications for some EA organizations that center around policy, although I think EA orgs are actually quite excellent on the metric of caring about local conditions. Here is the post in full.]

Confidence level: 85%

It has been said that all politics is local. I add that all progress is local.

I believe that there are some broadly applicable policies and structures that enable sustainable progress and growth. What economic progress works without property rights? Or what society obtains stable rule of law without decently high literacy rates? While property rights and literacy seem like no-brainers, the practice of progress requires more than identifying the directionally correct goals. It requires geographic reasoning. 

All things take place in a time and space, within a social network, under particular circumstances, which make them possible. The race is not always to the swift, nor insight to the wise, nor innovation to the bold, but time and chance affects them all. Or, in other words, social contingency is a powerful force.

If a handful of widely applicable rules for creating growth exist, I will not be surprised. Yet, I imagine that these rules, heuristics, or common structures will not be close to sufficient for securing the future of any people. Particular factors, unique to each situation are likely to swamp the causal force of whatever near-universal principles do exist. A leader’s ego, a heated political flareup, a bureaucrat’s charity or animus can be the difference between advancement or the indefinite tabling of an important initiative. 

Usually, fortune is not so personal and capricious that a single quirk in human personality can create and destroy many lives. But it can be. And people alter their behaviors in response to the personalities around them. 

Let’s take the public choice framework, my preferred method for thinking about the structures of institutions, the particulars of the immediate political settlement dominate the realm of the possible. Current incentive structures matter, and it is foolishness to think that biases, personalities, and soft networks (in addition to law, rules, and formal structures) do not create the incentive space. Hence the human geography of each environment matters very much. Triumphs and disasters are made within that geography.

Let me provide a case study of what I mean. Sri Lanka lost its sense and screwed itself royally. It began in May 2021, when the government announced it was going to shift to organic farming. On April 29 the ruling party passed a bill through parliament banning the import of chemical fertilizers. Since Sri Lanka has year round planting seasons, the effects set in quickly. The tea crop which makes up 11% of their exports plummeted by ~40%. By September the crop contractions made international news and were featured on Marginal Revolution. In May 2022, the prime minister was forced to step down amid protests and military deployment, the population has been cutting back on meals. Not a mere disruption. Sri Lanka has become a humanitarian crisis.

In none of its six public comments on Sri Lanka since the passing of the law has the State Department commented on the absolutely disastrous agricultural policy. Yet State does comment frequently about the ethnic tensions and violence that result at least partially from this mismanagement. Now down in the trenches, I am certain a State Department official or two were in contact with Sri Lankan counterparts throughout the legislative and post-legislative rollout of the law. Yet what good does that do?

In a more highly attuned world, some organizations would have identified the ill-fated law at its inception and marshaled resources to prevent the disaster well before the economic fortunes totally collapsed and politics devolved into violence and fear. Such an organization would be deeply involved in the country and have the reputation and weight and resources to act quickly and help the government backtrack faster, or even not pass such disastrous laws in the first place. Those marshaled resources would primarily be working relationships, phone calls, emails, personal meetings combined with scientific and economic expertise. 

By human geography and geographic reasoning I mean something particular. Human geography traditionally is the study of people and populations. Add into that, the study of local institutions, political settlement, and incentives and you get what I mean. Employing analysis of this type along with the more traditional and objective scientific and economic analyses keeps the innovator attuned to what is possible socially, legally, economically, and scientifically. Each level requires its own analysis.

Knowing in the abstract that zoning reform will be good for a city is not the same as being able to convince various interested parties in moving to a new equilibrium on zoning regulation. A separate set of social exigencies have set the equilibrium where it is. Adjusting the variables which created the current equilibrium opens up new social possibilities. If there is fear that the police force can’t handle the current level of density and prevent current murders, then that may be the factor preventing zoning reform from being a live political option. Being able to identify which changes will allow the targeted ‘high leverage’ changes to occur requires rich knowledge of the local terrain.

I think we should beware the temptation to see progress as an impartial set of principled actions. It can be naïve, self-defeating, even disastrous. 

The future is unequally distributed and so too are the tools of progress. Political ability might be clumped in an area where it is ineffective. Scientific ability might be wasted on the wrong problems. Economic understanding of the situation might be locked in the head of a few quiet bureaucrats. Identifying where and how a particular vista of progress can be opened is not merely a matter of skills, but finding the opportune places to deploy them, and thus the finding leverage in the right human geographies.