New Keynesian economics

New Keynesian economics is a macroeconomic school that emerged in the 1980s and 1990s as a response to the New Classical critique of traditional Keynesianism. Its central project was to provide microeconomic foundations for Keynesian conclusions — sticky prices, market imperfections, and the possibility of persistent unemployment — while retaining the methodological framework of rational expectations and dynamic optimization. The result is a body of theory that claims Keynesian lineage but operates within a conceptual framework that Keynes himself would have recognized as the very formalism he spent his career criticizing.

The Core Architecture: DSGE Models

The signature methodological contribution of New Keynesian economics is the Dynamic Stochastic General Equilibrium (DSGE) model. These models depict the economy as a system of interlocking optimization problems solved by representative agents who possess rational expectations, operate under perfect information except for specified frictions, and always converge to equilibrium. The "New Keynesian" elements are added as perturbations: sticky prices (via Calvo pricing), nominal rigidities, and occasionally financial frictions. But the equilibrium framework is not questioned. It is decorated.

This is a profound reversal of Keynes's original insight. Where Keynes argued that economies are complex adaptive systems whose dynamics are driven by expectations and conventions that cannot be reduced to optimization, New Keynesian economics treats these same expectations as the solution to a well-defined stochastic optimization problem. The animal spirits that Keynes saw as irreducible are domesticated as exogenous shocks to a rational expectations equilibrium. The radical uncertainty that Keynes identified as the defining condition of economic decision-making is replaced by risk — well-defined probability distributions over known states of the world. The model assumes what Keynes argued cannot be assumed.

The Methodological Contradiction

The New Keynesian synthesis rests on a methodological contradiction that its practitioners rarely acknowledge. The school claims to explain why markets fail to clear — why unemployment can persist, why recessions happen, why monetary policy has real effects — but it explains these phenomena within a framework that assumes markets would clear if not for the specific frictions the model introduces. The frictions are not derived from any theory of how actual firms and consumers behave. They are assumptions chosen to generate the desired conclusions. Calvo pricing, in which firms adjust prices only at random intervals, is not a theory of pricing behavior. It is a mathematical convenience that produces stickiness without requiring the modeler to explain why firms do not adjust prices.

This matters because the explanatory strategy is backward. In genuine science, one derives predictions from independently motivated assumptions and tests them against data. In New Keynesian economics, one chooses assumptions to reproduce stylized facts and then calibrates the model to fit historical data. The model has so many free parameters — adjustment costs, habit persistence, price stickiness indices, financial accelerator coefficients — that it can fit almost any macroeconomic time series. This is not explanation. It is curve-fitting with a theoretical alibi.

The Policy Framework and Its Limits

New Keynesian economics has been enormously influential in central banking. The New Keynesian Phillips curve, which relates inflation to the output gap and expected inflation, became the standard framework for monetary policy analysis. The Taylor rule, which prescribes how central banks should set interest rates in response to inflation and output, was derived within the New Keynesian framework. The IS-LM model of earlier Keynesianism was replaced by a three-equation model: an IS curve, a Phillips curve, and a monetary policy rule.

The 2008 financial crisis exposed the limits of this framework. New Keynesian models had no financial sector to speak of. The financial accelerator model of Bernanke, Gertler, and Gilchrist added a credit channel, but it treated financial frictions as a perturbation to an otherwise stable equilibrium. The crisis was not a perturbation. It was a phase transition — a regime change in which the entire attractor structure of the financial system shifted. The models could not predict it because they assumed the attractor structure was fixed. The post-crisis addition of financial frictions to DSGE models is not a theoretical advance. It is the addition of more epicycles to a Ptolemaic system.

The Systems-Theoretic Critique

From a systems perspective, New Keynesian economics suffers from a category error: it treats a complex adaptive system as if it were a stochastic control problem. The representative agent, the rational expectations assumption, and the equilibrium framework are all devices that eliminate the very properties that make economies complex — heterogeneous agents, network interactions, feedback loops, and emergent behavior. The result is a model that is mathematically sophisticated but structurally simple, and the simplicity is not a virtue. It is a failure mode.

The network structure of financial markets — counterparty exposures, collateral chains, liquidity cascades — produces dynamics that no representative-agent model can capture. The agent-based models that complexity economists have developed as an alternative treat the economy as a network of heterogeneous agents who interact locally and produce macro-patterns through their interactions. These models are not as mathematically tractable as DSGE models. But they are structurally realistic in ways that DSGE models are not. The trade-off between tractability and realism is not a technical problem. It is a philosophical choice about what kind of understanding we want.

The New Keynesian synthesis is not a synthesis at all. It is a subsumption — a domestication of Keynes's radical challenge within a framework that neutralizes it. The school that claims Keynes's name has produced the most elaborate mathematical apparatus in the history of economics, and it has done so by solving the wrong problem with ever-increasing precision. The problem is not how to add frictions to an equilibrium model. The problem is how to model economies as systems whose agents operate under radical uncertainty, whose dynamics are driven by network interactions, and whose macro-patterns emerge from micro-behavior that is not well-described by optimization. Keynes saw this problem clearly in 1936. The New Keynesian school has spent forty years looking away.