Sargent & Sims (2011): Untangling Cause and Effect in the Macroeconomy

The Hardest Question in Macroeconomics

Macroeconomics faces a problem that physics doesn’t: you can’t run controlled experiments on entire economies. You can’t raise interest rates in one copy of the United States while holding them steady in another, then compare the results. Everything happens at once — policy changes, oil shocks, technological shifts, consumer sentiment — and the economist must somehow untangle which caused what.

Thomas Sargent and Christopher Sims developed two complementary methods for solving this identification problem. Sargent showed how to use economic theory to interpret data when people’s expectations respond to policy changes. Sims showed how to let the data speak with minimal theoretical assumptions, tracing how shocks propagate through the economy. Together, they gave macroeconomists the tools to answer questions like: What happens to GDP and inflation when the central bank raises interest rates? How do economies adapt when governments change their fiscal rules?

Sargent: Structural Models and Rational Expectations

Thomas Sargent tackled the problem from the theory side. His approach: build explicit economic models where people form rational expectations about the future, then confront those models with data.

The challenge Sargent faced was profound. In the 1970s, Robert Lucas had shown that traditional econometric models were unreliable for policy analysis — because when policy changes, people change their behavior in response (the Lucas critique). A model estimated under one policy regime becomes useless for predicting what happens under a different regime.

Sargent’s solution: structural estimation with rational expectations.

• Build a model where households and firms optimize based on their expectations of future policy
• When policy changes, expectations change, and behavior changes — the model captures this feedback
• Estimate the model’s deep structural parameters — preferences, technology, information — that remain stable even when policy changes
• Use the estimated model to simulate counterfactual policy experiments

Key applications:

• The end of high inflation: Sargent studied how countries like Germany (1920s), Austria, Hungary, and Poland ended hyperinflation. His finding: credible, permanent changes in fiscal and monetary policy can stop inflation quickly — because rational people immediately adjust their expectations when they believe the policy change is real. This challenged the view that disinflation must be slow and painful
• The evolution of monetary policy: Sargent analyzed how the U.S. shifted from high-inflation policies in the 1970s to low-inflation policies in the 1980s, showing how policymakers themselves learned from experience — a process he modeled as Bayesian learning
• Fiscal policy and government debt: Sargent showed how expectations about future taxes and spending affect current economic behavior — the “unpleasant monetarist arithmetic” that links fiscal deficits to future inflation

Sims: Vector Autoregressions and Letting Data Speak

Christopher Sims approached the problem from the opposite direction. Instead of imposing detailed theoretical structure, he developed a method that requires minimal assumptions — the vector autoregression (VAR).

The insight: macroeconomic variables — GDP, inflation, interest rates, money supply — are all interconnected. Each variable depends on its own past values and the past values of all other variables. A VAR captures these interdependencies without requiring the economist to specify exactly how the economy works.

How VARs work:

• Collect time series data on key macroeconomic variables
• Estimate a system of equations where each variable is regressed on lagged values of all variables in the system
• Identify structural shocks — unexpected changes in monetary policy, oil prices, or technology — by imposing minimal restrictions on how shocks affect variables contemporaneously
• Trace out impulse response functions — how each variable responds over time to each type of shock

What Sims discovered:

• Monetary policy effects: When the central bank unexpectedly raises interest rates, output falls after a lag of several quarters, and inflation declines gradually. These impulse responses became the benchmark facts that any macroeconomic theory must match
• The price puzzle: Early VAR models showed that interest rate increases were followed by rising prices — the opposite of what theory predicts. Sims showed this was because the central bank raises rates in anticipation of inflation, not because rate increases cause inflation. Properly controlling for the central bank’s information resolved the puzzle
• Fiscal multipliers: VARs have been used to estimate how much GDP increases when the government increases spending — a question of enormous practical importance during recessions

Two Approaches, One Goal

Sargent and Sims represent two philosophies of empirical macroeconomics:

• Sargent’s structural approach: Start with economic theory, build a complete model, estimate its parameters, and use it for policy analysis. The advantage: you can simulate counterfactual policies. The risk: if the model is wrong, the conclusions are wrong
• Sims’s VAR approach: Start with data, impose minimal structure, and let statistical patterns reveal how the economy works. The advantage: you don’t need to get the theory exactly right. The risk: without theory, you may confuse correlation with causation

In practice, the two approaches have converged. Modern macroeconometrics uses structural VARs — combining Sims’s statistical framework with Sargent’s theoretical discipline. Central banks worldwide use both methods daily to guide monetary policy.

Their 2011 Nobel Prize was awarded “for their empirical research on cause and effect in the macroeconomy.”

Explain It to a Child

Imagine you’re trying to figure out what makes your plant grow. You water it, give it sunlight, add fertilizer, and talk to it — all at the same time. When it grows, which thing helped? That’s the problem Sargent and Sims solved for the economy. Sargent’s approach: build a model of how plants work — roots absorb water, leaves absorb sunlight — and use that model to figure out what matters. Sims’s approach: grow many plants with different combinations and carefully track what happens — without needing to understand exactly how roots work. One uses theory to understand data; the other uses data to discover patterns. Both figured out how to answer the big question: when the government changes something, what actually happens to jobs and prices?