New Neoclassical Synthesis

• New Neoclassical Synthesis is a macroeconomic paradigm that combines microfoundations with nominal rigidities to underpin DSGE modeling.
• It uses intertemporal optimization and rational expectations, integrating Calvo pricing and Taylor-rule monetary policies for empirical relevance.
• Its application in policy design highlights practical insights while facing critiques like observational equivalence and predictive limitations.

The New Neoclassical Synthesis (NNS) constitutes the dominant paradigm in modern macroeconomics, providing the intellectual and methodological foundation for the contemporary use of Dynamic Stochastic General Equilibrium (DSGE) models in both academic and central banking environments. The NNS is not a single theory but an overview that integrates the microfoundational rigor of New Classical economics with the nominal and real frictions introduced by New Keynesian economics. This synthesis has structured macroeconomic analysis since the 1990s, fundamentally influencing the formulation of policy-relevant models but also provoking debates over empirical adequacy, generality, and the evolution of the field’s methodology (Damiani, 1 Sep 2024, Chatelain et al., 2019).

1. Historical Development and Theoretical Foundations

The NNS emerged in the 1990s as a consensus framework, coinciding with a generational shift from the fragmented macroeconomic landscape of the 1980s—characterized by parallel developments in monetarism, real business cycle (RBC) theory, and various non-microfounded Keynesian models—to the methodological dominance of DSGE models. This synthesis was not the result of conclusive empirical superiority but rather a “tacit bargain” among leading researchers. The resulting intellectual settlement fused five elements:

1. Ramsey (intertemporal) optimization: Agents—including households and firms—solve explicit intertemporal problems.
2. Rational expectations: Agents form model-consistent forecasts, rendering anticipated policy actions neutral on real variables.
3. Walrasian general equilibrium: Market-clearing conditions, though subject to certain frictions.
4. Nominal rigidities (Calvo pricing): Nominal wage and price stickiness, introduced for empirical fit rather than fully derived microfoundations.
5. Taylor-rule monetary policy: Interest rate decisions respond to inflation and output, internalizing the time-inconsistency critique.

This convergence is frequently characterized as a “mutation” or “collage” rather than a Kuhnian scientific revolution, as it entailed selective adoption of previously mutually exclusive assumptions (e.g., microfoundations with sticky prices; monetary neutrality with policy effectiveness) (Chatelain et al., 2019).

2. Structural Features and Mathematical Frameworks

In NNS-based DSGE modeling, aggregate dynamics arise directly from agent-level intertemporal optimization subject to market frictions and policy constraints. The unifying mathematical structure typically incorporates:

• Household optimization:

$$\mathbb{E}_t \sum_{t=0}^\infty \beta^t [z_t \ln C_t + \ln(M_t/P_t) - (1/\eta) L_t^\eta]$$

• Firm behavior:
  • Finished goods firms operate under (possibly) perfect competition, using CES aggregation of intermediate goods.
  • Intermediate goods firms are monopolistically competitive, subject to Rotemberg or Calvo price adjustment costs.
• Market Equilibrium: Aggregate demand and supply linked through resource constraints and price setting.
• Monetary policy rules (Taylor-type): The nominal interest rate responds to deviations in inflation and output from target values.

Two canonical equations encapsulate NNS macro-dynamics:

• Expectational IS Curve:

$$\hat{x}_t = \mathbb{E}_t \hat{x}_{t+1} - (\hat{r}_t - \mathbb{E}_t \hat{\pi}_{t+1}) + \left[1 - \frac{1}{\eta}\right](1-\rho_z)q_t$$

• New Keynesian Phillips Curve:

$$\hat{\pi}_t = \beta \mathbb{E}_t \hat{\pi}_{t+1} + \frac{\eta}{\phi}(\theta-1) \hat{x}_t - \hat{e}_t$$

These expressions reflect forward-looking behavior, rational expectations, the persistent role of nominal rigidities, and the centrality of exogenous and endogenous shocks. The stability and solution of such models often depend on conditions for a unique equilibrium (e.g., the Blanchard-Kahn conditions) (Chatelain et al., 2019, Damiani, 1 Sep 2024).

3. The Synthesis: Resolving New Classical and New Keynesian Tensions

The NNS is predicated on the notion of complementarity between the New Classical and New Keynesian schools:

• From New Classical economics: Microfoundations (explicitly modeled optimization), dynamic general equilibrium structure, and rational expectations. These components ensured internal consistency and response to the Lucas critique, which warned that ad hoc models would produce misleading policy implications if parameter stability could not be maintained under regime changes.
• From New Keynesian economics: Nominal and real rigidities (sticky prices/wages, imperfect competition). By introducing adjustment costs and sectoral frictions, these models could rationalize persistent unemployment and the short-run nonneutrality of monetary policy, within a microfounded framework.

This synthesis allowed for both the demonstration that only unexpected policy shocks affect real activity in frictionless models (New Classical result), and that, given sufficient price or wage stickiness, systematic policy could exert real effects in the short run (New Keynesian extension).

The NNS thus yielded a modeling approach in which:

• All agents are forward-looking and optimizing.
• Nominal sluggishness or imperfect competition ensures departures from invariance to monetary policy.
• Both anticipated and unanticipated shocks matter, depending on the precise configuration of the model’s frictions.

4. Applications in DSGE Modeling and Policy Analysis

Post-synthesis DSGE models became the reference tool for both academic research and, critically, central bank policy design. Initial RBC-based DSGE models lacked monetary policy transmission and nominal rigidities. NNS innovation introduced:

• Monetary and fiscal sectors with explicit shock processes (technology, policy, preferences, cost-push, etc.).
• Model-based simulation of policy counterfactuals under various regimes of policy credibility and transmission, with explicit expectations and shocks.
• Calibrated or estimated structural parameters tailored to the economy-specific institutional context of each central bank: the era of “economy-specific DSGE tools.”

This development marked a methodological shift: the field moved away from the pursuit of universal, context-free explanatory models (as championed by Walras, Keynes, Friedman) toward a pragmatic focus on applied, context-sensitive analysis, acknowledging the heterogeneity of institutional and structural relationships across economies (Damiani, 1 Sep 2024).

5. Methodological Limitations and Critiques

Loss of Generality and Empirical Grounding

The NNS/DSGE paradigm is criticized for sacrificing the former aspiration to universal theory, as models are increasingly built for specific times and places, with context-dependent parameters and idiosyncratic shocks. The “loss of generality” is characterized as a central consequence of the synthesis (Damiani, 1 Sep 2024).

Microfoundations and Behavioral Assumptions

Recent empirical studies raise questions about the descriptive adequacy of rational optimization and representative agent assumptions. In response, DSGE models have been extended to include bounded rationality and endogenous heterogeneity, but the core NNS paradigm remains grounded in rational expectations and intertemporal optimization (Damiani, 1 Sep 2024).

Exogenous Shocks and Endogeneity

NNS-based models’ reliance on exogenous stochastic shocks (e.g., AR(1) processes for technology or policy) to drive macroeconomic fluctuations is subject to the critique that real-world phenomena are often endogenously determined, and hence not suitably modeled as exogenous, data-generating processes (Damiani, 1 Sep 2024, Chatelain et al., 2019).

Observational Equivalence and Identification

Multiple structural models—including those with rational or adaptive expectations, or alternative mechanisms of persistence—can be constructed to produce identical observable implications (observational equivalence). For example, for any linear rational expectations DSGE model, an observationally equivalent adaptive expectations VAR can be specified. Identification problems also persist, whereby deep parameters are not uniquely recoverable from the data, undermining the empirical content of the theory (Chatelain et al., 2019):

$$A_0(\theta) y_t = A_1(\theta) E_t (y_{t+1}) + A_2 y_{t-1} + A_3(\theta)x_{t} + u_t$$

Predictive Failures and Scientific Progress

The inability of standard NNS/DSGE models to anticipate major crises (such as the 2008 global financial crisis) has intensified scrutiny of their core assumptions. Moreover, the field is characterized as lacking cumulative scientific progress: models are modified or hybridized to accommodate new facts, rather than selected or rejected by decisive empirical tests. The paradigm is seen, in this respect, as sustained by methodological alliance rather than conclusive evidence (Chatelain et al., 2019).

6. Summary Table: Key Comparative Dimensions

Dimension	New Classical	New Keynesian	NNS/DSGE Synthesis
Microfoundations	Yes	Yes (add sticky prices/wages)	Yes
Expectations	Rational	Rational	Rational
Nominal Rigidities	No	Yes (wages, prices)	Yes
Market Structure	Perfect Competition	Imperfect Competition	Both, as appropriate
Policy Effectiveness	Only unanticipated	Present with rigidities	Present (short-run)
Core Model Type	DGE	DGE + Frictions	DSGE
Mathematical Tools	Dynamic programming	+ Calvo/Rotemberg pricing	+ Stochastic/dynamic opt.
Criticisms	Ignores frictions	Sometimes ad-hoc foundations	Loss of generality, empirical failures

7. Current Trajectory and Prospective Evolution

Ongoing critique and empirical challenges have led to efforts to further extend the NNS paradigm, notably by incorporating behavioral realism (bounded rationality, agent heterogeneity) and endogenizing sources of shocks. The future of macroeconomic modeling may thus entail a new synthesis that responds both to empirical requirements and to theoretical advances, guided by the limitations diagnosed in the NNS and the evolving demands of economic policy analysis (Damiani, 1 Sep 2024, Chatelain et al., 2019).

References to Key Papers

• [An essay on the history of DSGE models, (Damiani, 1 Sep 2024)]
• [Publish and Perish: Creative Destruction and Macroeconomic Theory, (Chatelain et al., 2019)]