Microfounded models

What Are Microfounded Models?

Microfounded models are a class of economic models in economic theory where macroeconomic relationships are derived from the optimizing behavior of individual economic agents, such as households and firms. Instead of positing direct relationships between aggregate variables, these models build from the ground up, linking collective phenomena to the underlying decision-making processes of micro-level entities. This approach aims to ensure internal consistency and robustness, particularly when analyzing the effects of policy changes. The development of microfounded models is a significant aspect of modern macroeconomics.

History and Origin

The concept of microfoundations emerged prominently in economics following World War II, as part of the "neoclassical synthesis" which sought to integrate neoclassical economics with Keynesian macroeconomic principles. While the notion of linking micro and macro principles had earlier roots, the "microfoundations project" gained significant traction in the 1970s. This was heavily influenced by economist Robert Lucas, Jr.'s critique of traditional macroeconometric forecasting models, which he argued were unreliable for predicting policy effects because their parameters were not invariant to changes in government policy¹².

Lucas asserted that the structure of econometric models, being based on the optimal decision rules of agents, would systematically alter when policy changed. This "Lucas Critique" highlighted the need for models grounded in the "deep parameters" of individual behavior, such as preferences and technology, which were assumed to be more stable¹¹. This intellectual shift led to a widespread demand for macroeconomic models to be explicitly derived from microeconomic analysis and optimizing behavior, a move detailed in academic discussions on the quest for microfoundations⁹, ¹⁰.

Key Takeaways

• Microfounded models derive macroeconomic outcomes from the optimizing behavior of individual agents.
• They aim to provide a consistent framework for economic analysis, particularly in response to policy changes.
• The "Lucas Critique" was a significant driver behind the adoption of microfoundations in macroeconomics.
• Dynamic Stochastic General Equilibrium (DSGE) models are a prominent example of microfounded models.
• These models are widely used by central banks for forecasting and monetary policy analysis.

Interpreting Microfounded Models

Interpreting microfounded models involves understanding that aggregate economic phenomena, such as inflation or unemployment, are outcomes of millions of individual choices and interactions. Unlike older, reduced-form macroeconomic models that might simply posit a relationship between, for instance, aggregate consumption and income, a microfounded model would derive the consumption behavior from the utility functions of individual households maximizing their lifetime well-being subject to budget constraints.

The insights gained from microfounded models are not typically a single number or indicator. Instead, interpretation focuses on how different policy interventions or external shocks propagate through the economy, affecting individual agents' decisions and, consequently, aggregate variables. For policymakers, this means understanding the behavioral channels through which their actions will impact the economy, rather than relying on empirical correlations that might break down under new policy regimes.

Hypothetical Example

Consider a central bank evaluating the impact of a potential interest rate change on aggregate investment. In a traditional macroeconomic model, the central bank might use a simple equation where investment is inversely related to the interest rate, based on historical data.

In a microfounded model, the analysis would be much more granular:

1. Households: The model would consider how the interest rate change affects households' saving decisions, which in turn influences the supply of loanable funds.
2. Firms: Firms, as economic agents, would optimize their investment decisions based on the cost of borrowing (influenced by the interest rate), expected future profits, and their production technology.
3. Interaction: The model would then simulate the interaction between these households and firms in a dynamic, general equilibrium setting.
4. Aggregate Outcome: The aggregate investment level would emerge as a result of these micro-level optimizing behaviors, providing a more detailed and consistent picture of the interest rate's overall effect, considering how agents adapt their behavior.

This approach allows economists to understand the underlying mechanisms rather than just the observed correlations.

Practical Applications

Microfounded models, particularly Dynamic Stochastic General Equilibrium (DSGE) models, have become a standard tool in modern macroeconomics. Their applications are widespread, especially in policy analysis and forecasting:

• Central Banking: Many central banks, including the Federal Reserve and the European Central Bank, utilize DSGE models to inform their economic outlooks, analyze the effects of monetary policy decisions, and conduct counterfactual simulations⁶, ⁷, ⁸. These models help policymakers understand how their interventions might influence aggregate demand, supply, and other key macroeconomic variables.
• Fiscal Policy Analysis: Governments and international organizations employ microfounded models to assess the impact of changes in fiscal policy, such as tax reforms or government spending programs, on economic growth, employment, and income distribution.
• Academic Research: These models are essential in academic economic research, serving as frameworks to test new theories, study business cycles, and analyze the propagation of various economic shocks.

Limitations and Criticisms

Despite their widespread adoption, microfounded models face several limitations and criticisms:

• Complexity and Calibration: Building and solving these models can be highly complex, requiring numerous assumptions and sophisticated econometrics. The models often rely on calibration, where some parameters are set based on existing empirical evidence rather than solely estimated, which can introduce subjectivity.
• "Representative Agent" Assumption: Many microfounded models, especially DSGE models, rely on the "representative agent" assumption, which assumes that the economy can be modeled as if it consists of a single, average household or firm. Critics argue this oversimplifies real-world heterogeneity among agents, which can be crucial for understanding phenomena like inequality or financial crises⁵.
• Forecasting Performance: While theoretically robust, some scholars argue that the forecast performance of DSGE models can be poor, particularly in predicting individual variables or capturing all the dynamics between macroeconomic time series⁴. They were criticized for not adequately predicting the 2008 financial crisis, leading to efforts to incorporate financial frictions more explicitly³.
• Limited Learning and Behavioral Aspects: Standard microfounded models often assume rational expectations and perfect information, which may not fully reflect how agents learn, adapt, or exhibit bounded rationality in real-world scenarios. Some critiques suggest that these models may not fully address the Lucas Critique if deep parameters like risk aversion can also change with policy². A detailed discussion on the dangers in the microfoundations consensus further elaborates on these points¹.

Microfounded Models vs. Dynamic Stochastic General Equilibrium (DSGE) Models

While closely related, it's important to distinguish between microfounded models as a general approach and Dynamic Stochastic General Equilibrium (DSGE) models as a specific type within that approach.

• Microfounded Models (General Concept): This term refers to any economic model where aggregate relationships are explicitly derived from the optimizing behavior of individual agents, rather than simply asserted. The core idea is to build macro from micro.
• DSGE Models (Specific Application): DSGE models are a dominant class of microfounded models that gained prominence in macroeconomics. They are characterized by their dynamic nature (accounting for intertemporal decisions), stochastic shocks (incorporating uncertainty), and the use of general equilibrium analysis (modeling the simultaneous interaction of all markets and agents).

Essentially, all DSGE models are microfounded models, but not all microfounded models are necessarily DSGE models in their full, modern form, although the terms are often used interchangeably due to the prevalence of DSGE models in modern macroeconomic policy and research. Confusion arises because DSGE models represent the most common and advanced application of the microfoundations principle in mainstream macroeconomics.

FAQs

Why are microfoundations considered important in macroeconomics?

Microfoundations are important because they aim to make macroeconomic models more robust and consistent. By deriving aggregate relationships from individual behavior, these models are designed to better predict the effects of policy changes, as they account for how individuals might alter their actions in response to new policies. This addresses the "Lucas Critique," which highlighted that relationships observed from historical data might not hold true when policy regimes change.

How do microfounded models differ from traditional aggregate models?

Traditional aggregate models often posit direct relationships between macroeconomic variables (e.g., aggregate consumption depends on aggregate income) without explicitly modeling the underlying individual choices. Microfounded models, conversely, start with individual utility maximization and profit maximization problems, and then aggregate these individual behaviors to derive the macroeconomic relationships. This provides a deeper, behavioral underpinning for the aggregate dynamics.

Are microfounded models always accurate in their predictions?

No, like all forecasting models, microfounded models have limitations and are not always perfectly accurate in their predictions. They rely on simplifying assumptions, such as the representative agent, and may struggle to capture complex real-world dynamics like financial crises or significant heterogeneity among economic agents. Continuous research aims to improve their predictive power and incorporate more realistic features.

What is the "representative agent" assumption in microfounded models?

The "representative agent" assumption simplifies modeling by assuming that the entire economy can be represented by a single, hypothetical individual or firm whose optimizing behavior reflects the average behavior of all agents. While this makes models more tractable, it is a significant simplification that critics argue can obscure important insights related to distribution, inequality, or diverse responses to economic shocks.

Do central banks use microfounded models for policy?

Yes, many central banks worldwide widely use microfounded models, particularly Dynamic Stochastic General Equilibrium (DSGE) models, as a key tool for policy analysis and forecasting. These models help them understand the potential effects of monetary policy actions, such as interest rate adjustments, on various economic indicators, and to assess different policy scenarios.