Skip to content

Causal Reasoning

How to reason about statements like “The US Develop Because of Tarrifs?’ Use more Narrow / Fine Grain Claims like; In Order to Develop Industry X; Wee Need Some Help; Until the Industry Become Competitive.

Is there evidence that protection (tariffs) helped industries achieve minimum efficient scale (MES) faster than without it? Which instances required tarrifs? See the Survivorship Bias.

Difference-in-Differences (DiD) is a statistical technique used in econometrics and other social sciences to estimate causal relationships when randomized controlled trials are not possible. It is particularly useful in situations where researchers want to measure the impact of a treatment or intervention (e.g., a policy change, new law, or economic shock) on a particular group or industry, while controlling for other factors that might affect the outcome.

So "tariffs are good" or "tariffs are bad" is a hopelessly broad question. Under what precise conditions do moderate tariffs in the X sector help a country at Y stage of development? Can a complex industry in which there is external competition be develop without any policy intervention?

Tariffs are instruments, not magic bullets.

A graph showing that "nations with higher tariffs grow slower" does not prove that tariffs are bad, because:

  1. Correlation is not causation:

    Maybe slower-growing countries happen to use tariffs more, because they are trying to protect weak industries, not that tariffs caused the slow growth.

  2. Confounding variables:

    Many other factors could be influencing both tariffs and growth — like political instability, weak institutions, bad infrastructure, bad education systems, etc.

  3. Reverse causality:

    It's also possible that countries that are already doing poorly resort to tariffs as a defensive move. (Bad growth → Tariffs, not Tariffs → Bad growth.)

  4. Selection bias:

    If you look at only certain countries (for example, poor ones that use tariffs), you might be missing countries where tariffs actually helped (selective survival).

  5. Measurement problems:

    "Tariff level" is very crude — different industries, different sectors, different moments in development need different levels of protection. A single number like "average tariff" misses the detail.

Causality

  • Key idea: Understanding and identifying cause-and-effect relationships between economic variables rather than mere correlations.
  • Modeling focus: Establishing the direction and magnitude of causal impacts for policy evaluation, forecasting, and theory testing.
  • Tools and methods:
  • Econometric techniques: Instrumental variables (IV), difference-in-differences (DiD), regression discontinuity design (RDD), propensity score matching.
  • Structural causal models: Directed acyclic graphs (DAGs), potential outcomes framework.
  • Natural experiments: Exploit exogenous variation to infer causality.
  • Counterfactual analysis: Assessing what would happen under alternative scenarios.
  • Importance: Crucial for credible policy analysis and understanding mechanisms behind economic phenomena.
  • Applications: Evaluating the effect of education on earnings, tax policy impacts, labor market interventions, health economics.

References

  • https://pmc.ncbi.nlm.nih.gov/articles/PMC7255316/
  • …