Software as an Institution

The rules of the game, known to economists as institutions and to managers as corporate culture, usually entail inoperable ideas. That is, any country or business has some rules, but these rules coincide neither with optimal rules nor with leadership vision. Maybe with an exception of the top decile of performers or something like this.

This inoperability isn’t surprising since the rules have obscure formulations. Douglass North and his devotees did best at narrowing what “good institutions” are, but with North’s bird-eye view, you also need an ant-eye view on how changes happen.

An insider perspective had been there all the time, of course. Organizational psychology and operations management organized many informalities happening in firms. In general, we do know something about what managers should and shouldn’t do. Still, many findings aren’t robust as we’d like them to be. There’s also a communication problem between researchers and practitioners, meaning neither of the two cares what the other is doing.

These three problems—formulation, coverage, and communication of effective rules—have an unexpected solution in software. How comes? Software defines the rules.

Perhaps Excel doesn’t create such an impression, but social networks illustrate this case best. After the 90s, software engineers and designers became more involved in the social aspects of their products. Twitter made public communications shorter and arguably more efficient. In contrast to anonymous communities of the early 2000s, Facebook insisted on real identities and secure environment. Instagram and Pinterest focused users on sharing images. All major social networks introduced upvotes and shares for content ranking.

Governance in online communities can explain success of StackExchange and Quora in the Q&A space, where Google and Amazon failed. Like Wikipedia, these services combined successful incentive mechanisms with community-led monitoring. This monitoring helped dealing with low-quality content that would dominate if these services simply grew the user base, as previous contenders tried.

Wikipedia has 120,000 active editors, which is about twice as many employees as Google has (or alternatively, twelve Facebooks). And the users under the jurisdiction of major social networks:


So software defines the rules that several billion people follow daily. But unlike soft institutions, the rules engraved in code are very precise. Much more so than institutional ratings for countries or corporate culture leaflets for employees. Code-based rules also imply enforcement (“fill in all fields marked with ‘*'”). Less another big issue.

Software captures the data related to the impact of rules on performance. For example, Khan Academy extensively uses performance tracking to design the exercises that students are more likely to complete — something that schools with all the experienced teachers do mostly through compulsion.

Finally, communication between researchers and practitioners becomes less relevant because critical decisions get made at the R&D stage. Researchers don’t have to annoy managers in trenches because software already contains the best practices. Like at that employed algorithms to grant its employees access privileges based on the past performance.

These advantages make effective reproducible institutions available to communities and businesses. That is, no more obscure books, reports, and blog posts about best practices and good institutions. Just a product that does specific things, backed by robust research.

What would that be? SaaI: software as an institution?

Public Policies and Persistence of Institutions

A fabulous series of works on institutional persistence emerged in the 2000s. To name a few:

This series has several things in common. First, the common narrative says that very old decisions influence current economic performance. Melissa Dell finds a 25% decline in current consumption due to forced labor in the 16-century mines. Nathan Nunn quantifies the impact of slavery on current output per capita in Africa. Abhijit Banerjee and Lakshmi Iyer report a 15% increase in current crop yields in territories owned by cultivators in the 19th century, compared to the territories of then-landlords:

Banerjee and Iyer, “History, Institutions, and Economic Performance.”
Banerjee and Iyer, “History, Institutions, and Economic Performance.”

A passerby may say that, of course, slavery and ownership last — that’s our history. This position is so general that it’s always true. The authors did much more than that. They showed by how much history matters — and the numbers are serious.

One unintended consequence, though. These results reassure the pessimists in developing countries. The problem is, of course, that developing countries aren’t developing much. The countries have some economic growth — often induced by commodity prices and imported technologies — and few successful fundamental reforms. People out there rarely see changes and don’t ask for them, so the past matters because it actually equals the present.

Let’s see how economists can encourage these people.

In a sense, any persistent connection between the past and the present is a public policy failure. If a 18th century earthquake destroyed a bridge and the local community didn’t rebuild it since then, the earthquake naturally worsens the current village performance. This happens due to the earthquake and the (absent) mitigation policies.

The earthquake stands for any lasting factor of development. A lasting factor (earthquake) determines policies (endure the loss of the bridge) and these policies affect current outcomes (village output). This is the standard line in the literature. The lasting factor can be anything. Institutions just happened to be the group large enough to be statistically significant in small samples, which always constrain research in economic growth. This group is a big bunch of laws and rules, which wouldn’t manifest itself in regressions if taken separately.

What do we know about such lasting factors?

Well, sometimes they last. Of course, there’s a publication bias favoring historical persistence. Imagine, instead, the flow of publications enumerating historical factors that do not affect the present. A lot of boring reading, for sure. Instead, we have a few dozens of highly cited papers with robust positive results. These citations signal two things. First, authors brilliantly did their job. Second, few results of this sort exist. Perhaps, citations would be diluted if results could be tested in different settings. Just look at randomized trials, which are very important but authors basically no longer seek publishing there as they test same policies across countries.

Secondly, historical factors leave many questions. Are they resistant to public policies? Are they resistant to informed public policies? To exogenous policies, like international aid programs? To a different set of institutions?

These questions hint at the idea that some societies respond to challenges and, therefore, flourish. But for the other societies, the data always shows persistence. Especially to large exogenous shocks, which dominate the literature because they are nearly the only way to measure the impact.

I mean, all the papers I mentioned exploit a very strong “treatment”: A well-armed army of foreigners attack an indigenous population and establish particular rules. Why is it common in publications? Because such a situation is accompanied by a suitable identification strategy: invaders grab whatever comes first (that’s randomization) and divide previously homogeneous regions (that’s discontinuity). Still, these invasions are only one part of history.

So we need to know more about the economic significance of this sort of persistence. We may discover that persistence vanishes. For example, the conflict between relatively equal France and Germany in the Napoleonic Wars (1803−1815):

The Consequences of Radical Reform: The French Revolution
The Consequences of Radical Reform: The French Revolution

The treatment group includes the German territories occupied by the French. The French induced growth-enhancing reforms, so the treatment group grows faster:

The Consequences of Radical Reform: The French Revolution
The Consequences of Radical Reform: The French Revolution

The difference is becoming unimportant as both groups develop. The control group reaches the same level of reforms with a 10-year lag, but it does reach the same level. The gap in urbanization is also closing relative to the level of urbanization.

History matters until you find solutions to very specific problems. Returning to Melissa Dell’s 25% consumption gap in Peru, what would happen to this gap if Peru got richer? What would happen with persistence in growing India and Africa, featured in the other articles? The good question is, therefore, what can we do about this persistence?

Macroeconomics Models and Force of Habit

The public rightly questioned macroeconomics and academic finance after the 2008 burst. Record housing prices and debt, both relative to income, look a plausible cause for concern and they are. Why, then no one prevented it?
The design of the markets discourages companies from being overly cautious. Banks didn’t quit inflated housing markets because these markets were still inflating. Profits reinforce participation.
The designers of the markets had got obvious signals too late to avoid consequences. And very few wanted to be the person who bursts balloons with a needle at a birthday party anyway. Governments and central banks waited for problems to come first.
Many more versions exist. But none of them can explain the bubble with lack of knowledge alone. People in finance see housing prices every day, and high ratios are quite telling, apart from answering the question, “When will this trend end?”
Designers and players played by the rules, and they certainly had selfish incentives. Academia was relatively free of these rules and incentives. Did macroeconomists have selfish incentives to find a bubble, instead?
Yes and no. You will barely find a major university economist who likes forecasting. Because sometimes the predictions come true. Thus, sometimes they don’t. Economists prefer discussing things that have happened already. And they do it unhurriedly. Operative policy interventions are unlikely in the environment where even publishing an academic paper takes up to several years.
More so, it’s difficult to find a serious academic paper that includes policy recommendations. Scholars explain things that have occurred. Policymakers can use these insights to forecast. By 2008, policymakers had models. Were these models good? They happened to have specific limitations. But even bankers had incentives to use the best models they might get to quit the housing market in time.
There’re no obstacles to adopting models with better predicting power. Then, maybe policymakers did use the best models they had? Rather, they used the most reliable equations: the ones that they understand and used for years. And DSGE models won over various alternatives, including those by heterodox economists, who offered equations that predicted the crisis.
A theory that predicts one-in-fifty-years events is not trusted because it can hardly earn a reputation of a reliable one. No, the theory itself may be predictive and great, but it lacks an empirical base to show its fitness. That makes this theory and underlying models an unlikely candidate for widespread use.
Macroeconomics is responsible for not knowing enough in the sense of biologists who don’t know how to cure cancer. There’re wrong turns and no malicious incentives. Right turns require outstanding efforts and time. Including time for gathering unique data, like the data that came from the terrible Great Recession. Bad theories still can be the best, until we have more evidences. Economics works when we recognize limitations of previous theories and try to build better ones.