Getting started

The confidence with which we express opinions about economics, politics, or climate should worry us. We all know a simple solution to a difficult problem. We want affordable housing, less hunger, a cleaner environment, and a more prosperous society. The intention may be honorable. That does not prove the chosen measure will achieve it.

In social affairs, every decision enters a network of people who make decisions of their own. A landlord, a worker, a business, a government official, or a family adjusts its behavior when the rules change. Others then react. The outcome appears at the end of that chain, sometimes long after the first decision. If we stop at the initial announcement, we leave out the decisive part of the story.

The paradox of partial knowledge

We know enough to have an opinion. Often, we still do not know enough to recognize its limits. That is the paradox of partial knowledge. It lets us take part in a conversation, but it can also lead us to mistake an appealing explanation for a complete one.

  • Neil deGrasse Tyson sums it up precisely: “One of the great challenges in life is knowing enough to believe you're right, but not enough to know you're wrong.”

The problem grows when we talk about economics, politics, or sociology. We all live within those systems. We all pay, work, vote, consume, or debate. Personal experience matters, but it offers a narrow window. A measure may ease the situation of the person in front of us and harm someone we cannot see. It may also work today and change the conditions that made it successful tomorrow.

Before defending a solution, it helps to distinguish four things. First, what we want to achieve. Second, what we say we want to achieve. Third, what other people expect us to do. Fourth, what actually happens. All four may align. They may also diverge completely. An impressive speech does not close that gap; neither does a noble intention.

Sometimes the mistake comes from haste. At other times, vanity, convenience, or the political incentive to promise immediate results plays a part. We do not have to accuse everyone of bad faith to ask a more serious question: what will others do when we do this? If we cannot answer, we should at least acknowledge that the outcome remains open.

What makes a system complex

A complex system consists of parts that interact continuously. Its behavior cannot be understood by adding up what each part would do on its own. Changing one part changes the relationships among the others. Some responses reinforce the change; others slow it down. This is what we call feedback.

Consider a couple. Each person interprets what the other does, responds, and remembers what they have experienced together. The same sentence may inspire trust one day and rejection another. Its effect depends on the relationship built over time, not just the words spoken. A couple is a small system. Its dynamics already show why counting the parts is not enough to understand the whole.

Something similar happens in an economy on a larger scale. Consumers, businesses, workers, banks, investors, and public authorities observe one another's decisions. Each has incomplete information and goals of their own. When a rule changes, nobody stands still waiting for the prediction of those who approved it to come true. They adjust prices, work, savings, investment, or consumption according to what they think is possible and worthwhile.

Those adjustments can repeat. A decision prompts a response; the response changes the original problem; someone intervenes again. That is why time matters. The first effect may be visible and easy to count. Later effects may appear in another city, another sector, or among people who were not part of the debate.

There are also thresholds. A small change may have no noticeable effect until other changes accumulate. Then the behavior of the whole system shifts rapidly. This possibility does not mean that every detail will trigger a catastrophe. It means relationships do not always follow a straight line.

A common confusion arises here. The so-called butterfly effect refers to systems in which a tiny initial variation can become greatly amplified. Complexity is a broader framework. It includes interaction, adaptation, networks, feedback, and outcomes that emerge from relationships among the parts. A system can be complex without every small variation having a huge effect. If we treat the two ideas as synonyms, we lose precision when we need it most.

Intention does not control the response

I want everyone to be able to access decent housing. That goal deserves attention. But if I propose a measure, I must ask how those who build, rent, buy, finance, or look for homes will respond. Will supply increase? Will it decrease? Will access conditions change? Who will receive help, and who will be left out? Unless I follow those responses, I cannot know whether my proposal puts housing within reach of more people or ends up closing doors.

The same applies to food aid. Providing food can meet an immediate need. That urgency deserves a response. But a prolonged intervention can also change the opportunities available to local farmers, producers, and merchants. The question is not whether to help or turn away. It is what help reaches people today and what capacity to produce will remain tomorrow. Poorly designed aid could harm local production or create dependency. It could also prevent a tragedy without causing those effects. We must study each case before turning an intuition into a universal prescription.

Environmental policies require the same care. If a measure reduces production in one place, we need to know whether demand disappears, shifts to another product, or moves that production elsewhere. A reduction recorded within one border could coexist with activity moving beyond it. That is a possibility we need to measure, not an outcome we can proclaim in advance. We must also count costs, benefits, technological alternatives, and effects on everyday life. Climate, industry, and human decisions are connected; a slogan does not calculate those relationships.

We can apply the same discipline to forest protection. A rule may aim to conserve a natural area and still change how it is managed. To judge it, we need to examine its effects on vegetation, water, people, and wildfire risk. Blaming a specific policy for a fire without that examination would repeat the very mistake we want to call out.

From everyday life to large networks

Complexity is not confined to economics.

  • During an epidemic, people change their behavior when they perceive a risk. That change affects their contacts and, in turn, the course of the problem they are trying to avoid.
  • In an ecosystem, species, nutrients, water, and temperature form relationships that affect one another.
  • In geopolitics, states, businesses, and alliances anticipate one another's responses before they act.
  • Financial markets offer another version of the same idea. Participants do not look only at a company or a price: they watch what they think others will do. Their expectations shape their trades, and those trades change the environment they were trying to anticipate.
  • In a society, ideas and behaviors circulate through families, institutions, media, and platforms. A public decision can change not only the resources available but also how people interpret what is acceptable or worthwhile.
  • Even physical and chemical systems show the limits of linear intuition. Turbulence, phase transitions, and certain oscillating reactions remind us that a system's behavior can change unexpectedly when its internal relationships change. We should not carelessly apply a physical law to politics.

We can, however, take a methodological lesson: when many parts interact, an explanation based on a single cause is often incomplete.

The economy is not a static board

Let us look at the people involved. People want well-being, security, and opportunity. Some study, work, save, or take risks to improve their situation. Others make different choices. There is no average citizen who fully represents everyone. The categories also overlap: a worker is also a consumer; a retiree may invest; someone in government also responds to incentives.

  • Workers seek stable employment and pay that compensates them for their time and effort.
  • Businesses need resources to produce and customers to sustain their operations.
  • Investors commit assets that they could spend or keep outside a business venture. That investment may finance products, services, and jobs. It also involves risk.

Reducing investors to a caricature keeps us from understanding what they will do if expected returns or the rules of the game change.

Government officials and those who hope to govern occupy a different position.

  • They can change rules, taxes, and public spending.
  • They must answer for their decisions, but they also seek support and face pressure.
  • A promise that offers an immediate benefit may be politically attractive even if its costs emerge later or fall on people who are less visible.
  • We should examine that incentive without assuming that every measure aims to benefit those in power.

Suppose a tax is imposed on a company.

  • The law may clearly identify who must remit it.
  • That still does not explain how the company, its customers, its workers, and its owners will respond.
  • Prices, wages, investment, or profits could change to different degrees.

The response will depend on the alternatives available to each actor and the adjustments that follow. Asking who sends in the payment is necessary. Asking who ultimately bears the cost opens up the real inquiry.

Think before making promises

Nobody can know all the consequences of a complex decision in advance. That uncertainty does not require inaction. It requires better work.

  • First, describe the problem without confusing it with the solution we prefer.
  • Then identify the actors and ask what incentives each has to adapt.

We need to look at immediate effects and those that might emerge after several rounds of response. And when we say that a measure caused an outcome, we must seek evidence and alternative explanations.

Perhaps the most useful question is the least comfortable: what would have to happen to prove me wrong? It protects us from enthusiasm for our own idea and from the ease with which we blame every failure on others.

In complex systems, intellectual humility does not diminish ambition. It allows us to correct course before the costs spread.

This article opens up a framework for analysis. In later articles, we can apply it to specific decisions. One of them has a familiar name:

  • the corporate income tax. Is it enough to point to the company? Do its shareholders bear the cost? Are we sure?

To answer, we will have to follow all the actors, their incentives, and their responses. That is where the work left undone by a slogan begins.