{
  "id": "new-neoclassical-synthesis",
  "title": "New Neoclassical Synthesis",
  "lang": "en",
  "source_lang": "en",
  "summary": "The New Neoclassical Synthesis combines optimizing, forward-looking macroeconomic models with nominal rigidities to explain economic fluctuations and analyze stabilization policy.",
  "category": "economics",
  "subcategory": "macroeconomics",
  "type": "concept",
  "aliases": [
    "New neoclassical synthesis",
    "new synthesis in macroeconomics"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/new-neoclassical-synthesis/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/new-neoclassical-synthesis/",
      "json": "https://aiwiki.page/zh/new-neoclassical-synthesis.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "The New Neoclassical Synthesis is an approach to macroeconomics that combines the intertemporal optimization and equilibrium methods associated with new classical and real business cycle research with the imperfect competition and nominal rigidities emphasized by New Keynesian economics. Its central proposition is that real forces shape productive capacity and long-run economic activity, while slow adjustment of prices and wages allows monetary disturbances and changes in aggregate demand to affect production and employment in the short run.\n\nThe expression became prominent through Marvin Goodfriend and Robert G. King’s 1997 essay, “The New Neoclassical Synthesis and the Role of Monetary Policy.” It describes a convergence in modeling methods and monetary analysis rather than a single model or a universally accepted doctrine. Its characteristic framework is a dynamic stochastic general equilibrium model in which households and firms make forward-looking decisions, markets interact, and some prices or wages cannot adjust immediately.\n\nThe synthesis supplied an influential language for studying inflation, business cycles, and central bank policy from the 1990s onward. Its simplest versions explain why credible inflation stabilization can improve economic performance, but their conclusions depend on assumptions about market distortions, financial arrangements, information, and policy constraints. Subsequent research has expanded the framework to incorporate financial crises, borrowing limits, household heterogeneity, and unconventional policy instruments."
    },
    {
      "heading": "Historical origins and the meaning of synthesis",
      "text": "The earlier neoclassical synthesis, associated especially with Paul Samuelson and postwar textbook economics, combined Keynesian explanations of deficient demand with neoclassical theories of allocation and growth. In practice, short-run analysis often used the IS–LM model, while long-run analysis relied on market-clearing models. These components were not always derived from the same assumptions about individual behavior. The newer synthesis sought a more unified structure: short-run non-neutrality and longer-run adjustment were to emerge within one explicitly specified intertemporal economy.\n\nTwo developments made that reconstruction important. First, the inflation experience of the 1970s weakened confidence in policy models that treated historical relationships between inflation and unemployment as stable. Second, Robert Lucas’s 1976 policy-evaluation critique argued that estimated behavioral relationships could change when policy rules changed. If households and firms understood a new policy regime, their expectations and decisions would adjust. A model intended to compare policy regimes therefore needed to explain behavior in terms of preferences, technology, constraints, and expectations rather than rely only on correlations observed under an earlier regime.\n\nDuring the 1980s, real business cycle theory, particularly the work of Finn Kydland and Edward Prescott, advanced quantitative models built around optimizing agents, stochastic technology, and equilibrium resource allocation. New Keynesian researchers developed explanations of why prices and wages might adjust slowly even when private agents pursued their own interests. Important contributions included John Taylor’s staggered-contract models, Guillermo Calvo’s price-setting framework, and models connecting imperfect competition to the output effects of nominal disturbances. These research programs disagreed about important mechanisms, but increasingly shared mathematical tools and an emphasis on explicit behavioral foundations.\n\nGoodfriend and King’s 1997 formulation identified the emerging combination of intertemporal optimization, rational expectations, imperfect competition, and costly price adjustment. Richard Clarida, Jordi Galí, and Mark Gertler’s 1999 survey, “The Science of Monetary Policy: A New Keynesian Perspective,” presented a closely related monetary-policy framework. Michael Woodford’s 2003 book, Interest and Prices, developed its theoretical foundations in detail. The label “New Neoclassical Synthesis” is therefore broader than any one contribution: it denotes an overlapping research program whose practical core became modern New Keynesian monetary economics."
    },
    {
      "heading": "Optimization, expectations, and equilibrium foundations",
      "text": "The synthesis begins with decisions made over time. A household chooses consumption, saving, and labor supply subject to its budget constraint. A representative specification is\n\n\\[E_0\\sum_{t=0}^{\\infty}\\beta^t \\left[ \\frac{C_t^{1-\\sigma}-1}{1-\\sigma} -\\chi\\frac{N_t^{1+\\varphi}}{1+\\varphi} \\right],\\]\n\nwhere \\(C_t\\) is consumption, \\(N_t\\) is labor, \\(\\beta\\) is a discount factor, and the remaining parameters govern preferences. The logarithmic consumption case is obtained as \\(\\sigma\\) approaches one. This formulation makes current spending depend on expected future income and returns, not simply on current disposable income.\n\nThe household’s Euler equation links the marginal utility of consumption today to the expected marginal utility of consumption tomorrow, adjusted for the real return on saving. Its labor-supply condition links the real wage to the marginal rate of substitution between consumption and leisure. Firms choose inputs to minimize costs and set prices to maximize expected discounted profits. These conditions connect aggregate behavior to microeconomics, although the use of a representative household does not by itself establish that aggregation accurately describes a population of different households.\n\nExpectations are usually modeled as rational expectations: agents’ forecasts are consistent with the probability distributions generated by the model and the information available to them. This does not mean that agents predict every event correctly. Shocks remain uncertain, and forecast errors occur. It means that agents do not systematically ignore information the model assumes they possess. Announced future policy can consequently affect current consumption, investment, and price setting before the policy instrument actually changes.\n\nEquilibrium requires that individual plans be mutually consistent with resource constraints and market conditions. Goods production must be allocated among consumption, investment, government purchases, and other specified uses. Asset positions must satisfy financial clearing conditions, and firms’ demand for labor must be consistent with household behavior. Unlike a perfectly competitive flexible-price model, however, equilibrium need not be efficient. Firms may charge markups, some prices may be inherited from earlier periods, and policy may influence the resulting allocation.\n\nThe synthesis thus separates two ideas sometimes conflated in discussion. Equilibrium means consistency among decisions and constraints; it does not necessarily mean full employment, an optimal allocation, or instantaneous adjustment. A recession can be an equilibrium response to a shock when prices are sticky and policy does not offset the resulting decline in spending. Nevertheless, basic models often abstract from explicit job search and involuntary unemployment, so their employment fluctuations should not automatically be interpreted as a complete theory of the labor market."
    },
    {
      "heading": "Imperfect competition and nominal rigidity",
      "text": "Monopolistic competition gives firms a reason to choose prices rather than take them as externally given. In a common specification, households purchase differentiated goods that are imperfect substitutes. Each producer faces a downward-sloping demand curve and charges a markup over marginal cost. With a constant elasticity of substitution \\(\\varepsilon>1\\), the desired flexible-price markup is \\(\\varepsilon/(\\varepsilon-1)\\). The resulting market power matters because production can remain below the efficient level even when every price is flexible.\n\nNominal rigidity is often introduced through Calvo pricing, based on Guillermo Calvo’s 1983 model. In each period, a firm can reset its price with probability \\(1-\\theta\\), independently of how long its existing price has been in place. A fraction \\(\\theta\\) cannot reset. The expected duration of a price spell is \\(1/(1-\\theta)\\) periods; for example, \\(\\theta=0.75\\) implies an average duration of four quarters in a quarterly model. This is a property of the assumed adjustment process, not a universal empirical estimate.\n\nA firm that can reset its price considers the period for which that price may remain effective. Its chosen price therefore depends on expected future costs, demand, and discount factors, not just on current conditions. Changes in inflation expectations can affect current inflation through this forward-looking decision. If nominal spending rises while many firms leave prices unchanged, demand shifts toward those firms and aggregate production can expand. Monetary policy is therefore not neutral in the short run.\n\nPrice dispersion is another important consequence. When some firms reset and others do not, relative prices differ for reasons unrelated to productivity or consumer preferences. Buyers redirect spending, and otherwise similar producers operate at different scales. With identical production technologies, this dispersion can raise the aggregate labor or input requirement needed to produce a given quantity of final consumption. Inflation therefore has a resource-allocation cost in addition to any inconvenience associated with changing a unit of account.\n\nCalvo pricing is analytically convenient but is not the only adjustment mechanism. Taylor contracts specify predetermined contract lengths; Rotemberg adjustment costs make rapid price changes costly; state-dependent models allow firms to adjust when the benefit exceeds a fixed cost. Wage rigidity can be modeled similarly and helps account for persistent movements in real wages and employment. These alternatives can generate different responses to large shocks, especially when high inflation changes firms’ incentives to adjust. The synthesis does not establish that one adjustment mechanism applies in every country or period."
    },
    {
      "heading": "The canonical three-equation model",
      "text": "A compact representation of the synthesis consists of an intertemporal demand equation, a New Keynesian Phillips curve, and a monetary-policy rule. The model is generally derived by approximating a more detailed economy around a steady state with low or zero inflation. Its central activity variable is the output gap, commonly written \\(x_t=y_t-y_t^n\\), where output is measured in logarithms and \\(y_t^n\\) is the flexible-price, or natural, level of output. This benchmark is model-specific and is not directly observable.\n\nThe demand relationship is\n\n\\[x_t=E_t x_{t+1} -\\frac{1}{\\sigma} \\left(i_t-E_t\\pi_{t+1}-r_t^n\\right).\\]\n\nHere \\(i_t\\) is the nominal short-term interest rate, \\(\\pi_{t+1}\\) is next period’s inflation, and \\(r_t^n\\) is the natural real interest rate consistent with flexible-price equilibrium. When the expected real interest rate, \\(i_t-E_t\\pi_{t+1}\\), exceeds the natural rate, households have an incentive to postpone consumption, reducing current demand relative to its flexible-price benchmark. Expected future activity also matters because consumption decisions are intertemporal.\n\nThe supply relationship is\n\n\\[\\pi_t=\\beta E_t\\pi_{t+1}+\\kappa x_t+u_t,\\]\n\nwhere \\(\\kappa>0\\) measures inflation’s sensitivity to the output gap and \\(u_t\\) represents a cost-push disturbance. In the simplest Calvo model, inflation is initially related to real marginal cost. Writing it in terms of the output gap requires additional assumptions connecting marginal cost to activity. With linear production and standard preferences, a familiar special case gives\n\n\\[\\kappa=\\frac{(1-\\theta)(1-\\beta\\theta)}{\\theta}(\\sigma+\\varphi).\\]\n\nOther technologies, wage-setting arrangements, and firm-specific factors change this expression.\n\nA stylized Taylor rule closes the model:\n\n\\[i_t=r_t^n+\\pi^* +\\phi_\\pi(\\pi_t-\\pi^*) +\\phi_x x_t,\\]\n\nwhere \\(\\pi^*\\) is the inflation target. Empirical rules often replace the time-varying natural rate with a constant intercept and include interest-rate smoothing or a monetary-policy shock. Variables and rates must be expressed in consistent units and relative to compatible steady states.\n\nThese equations show why the synthesis cannot be reduced to a mechanical relationship between money growth and output. Policy works through actual and expected real interest rates, while inflation depends on expectations and marginal costs. The system is forward-looking: a complete solution requires assumptions about shocks, the policy regime, and admissible equilibrium paths. The three-equation version usually omits physical capital, banks, international trade, and detailed fiscal arrangements; its clarity comes partly from those omissions."
    },
    {
      "heading": "Monetary-policy rules, credibility, and determinacy",
      "text": "The central policy implication is that monetary policy can improve short-run outcomes by stabilizing demand around a suitably defined benchmark. A central bank changes its short-term policy rate, affecting expected real financing costs and the timing of spending. In larger models, these effects pass through investment, exchange rates, asset prices, and borrowing conditions. The monetary transmission mechanism therefore depends both on current instrument settings and on beliefs about their future path.\n\nA crucial distinction concerns whether a policy rule produces a unique bounded equilibrium. In the simplest model, a sufficiently strong response to inflation helps prevent self-fulfilling fluctuations. The familiar Taylor principle requires the nominal interest rate to rise by more than one percentage point in response to a persistent one-percentage-point increase in inflation, other things equal. The expected real interest rate then rises rather than falls, restraining demand. With particular timing conventions, output responses, or additional economic mechanisms, the exact determinacy conditions differ.\n\nWithout an adequate response, an expectation of higher inflation can reduce the expected real rate, stimulate spending, and generate the inflation initially anticipated. Such a path need not be driven by a change in technology or preferences. This possibility explains why the synthesis emphasizes the policy regime rather than only individual policy decisions. Credibility is not simply a favorable reputation: it concerns whether private expectations are anchored by a coherent, understood, and feasible policy strategy.\n\nCommitment can improve stabilization when firms and households are forward-looking. A central bank that can credibly describe its future responses influences current price setting and consumption. Under discretion, by contrast, it reoptimizes period by period, potentially creating incentives to abandon earlier promises. Optimal commitment policies may display history dependence: current actions respond partly to earlier conditions or unfulfilled objectives, not only to the current inflation rate and output gap.\n\nThe framework helped provide theoretical support for inflation targeting, transparent communication, and analysis of central bank independence, but it does not uniquely prescribe a particular institutional system. Nor does it imply that a central bank must mechanically follow a numerical Taylor rule. Instrument rules, forecast-based procedures, and target criteria can implement related stabilization objectives. Their performance depends on information quality, model uncertainty, fiscal arrangements, and the constraints facing policymakers. Announcements also have limited influence when the public doubts that future actions will match them."
    },
    {
      "heading": "Welfare, efficient output, and stabilization trade-offs",
      "text": "A distinctive ambition of the synthesis is to derive policy objectives from household welfare rather than assume an arbitrary preference for smooth output and low inflation. Under restrictive conditions, a second-order approximation to utility yields a quadratic loss resembling\n\n\\[L=\\frac{1}{2}E_0\\sum_{t=0}^{\\infty}\\beta^t \\left(\\pi_t^2+\\lambda x_t^2\\right).\\]\n\nThe parameter \\(\\lambda\\) depends on structural features such as preferences, technology, and the degree of price rigidity. This expression is not universally valid: its derivation requires attention to the steady state, existing distortions, the definition of the output gap, and the order of approximation.\n\nInflation enters welfare because staggered price adjustment creates inefficient relative-price dispersion. An activity gap enters because production and labor effort can depart from their efficient levels. The efficient allocation must, however, be distinguished from the natural allocation. Natural output is what a model generates with flexible nominal prices while retaining its other distortions. Efficient output is the allocation that maximizes the relevant welfare criterion subject to resource and technological constraints. Monopolistic markups, distortionary taxes, or labor-market imperfections can separate the two.\n\nIn a particularly simple model, the steady-state markup distortion is offset by an appropriate subsidy and there are no time-varying real distortions. Stabilizing inflation then also stabilizes the welfare-relevant output gap. Olivier Blanchard and Jordi Galí described this alignment as the divine coincidence. It explains the strong case for price stability in benchmark versions of the synthesis: a single stabilization objective can address both nominal dispersion and inefficient activity fluctuations.\n\nThe coincidence breaks down when shocks alter the gap between natural and efficient output. A change in desired wage markups or another cost-push disturbance can raise inflation even as efficient activity weakens. Returning inflation immediately to target may then require an additional contraction. With both sticky wages and sticky goods prices, stabilizing goods-price inflation alone generally cannot eliminate all nominal distortions. Optimal policy may instead balance wage inflation, price inflation, and activity.\n\nThis welfare analysis also explains why “stabilizing output” does not normally mean keeping output constant or forcing it onto an unchanging trend. A persistent improvement in productivity can raise efficient production, while changes in preferences or resource availability can lower it. Policy should not necessarily offset such movements. The practical difficulty is that the efficient and natural benchmarks are latent, model-dependent quantities. A welfare result established inside a particular model consequently does not translate automatically into an observable, operational policy target."
    },
    {
      "heading": "Quantitative models and empirical evaluation",
      "text": "The early three-equation framework was primarily a tool for organizing mechanisms. Quantitative applications added capital accumulation, investment adjustment costs, consumption habits, variable capital utilization, sticky wages, and several disturbances. These additions helped reproduce persistent movements in output, inflation, and interest rates that a minimal model often struggled to match. They also made policy experiments more realistic by distinguishing consumption, investment, labor input, and other components of gross domestic product.\n\nTwo influential contributions were Lawrence Christiano, Martin Eichenbaum, and Charles Evans’s 2005 study of nominal rigidities and monetary shocks, and Frank Smets and Rafael Wouters’s estimated models, including their 2003 euro-area and 2007 United States studies. These models combined optimizing behavior with multiple adjustment frictions. Their importance lay partly in showing how the synthesis could be used for quantitative estimation rather than remain only a small theoretical construction.\n\nBayesian estimation became a prominent method. Researchers specify prior distributions for parameters and combine them with the likelihood implied by the model and observed data. The resulting posterior distribution supports parameter estimates, uncertainty intervals, forecasts, and comparisons among specifications. Data are often related to the model through a state-space representation, allowing unobserved shocks and latent variables to be inferred. Central banks and international institutions adopted related models alongside other forecasting and policy-analysis tools; the European Central Bank’s New Area-Wide Model is one example.\n\nEmpirical fit does not by itself settle whether the mechanisms are correct. A model can explain observed persistence through habits, sticky information, persistent shocks, or other devices that imply different policy responses. Identification is difficult when several structural parameters generate similar movements in aggregate data. Results can also depend on prior distributions, detrending choices, the variables included, and the stability of the assumed policy regime.\n\nFor these reasons, evaluation extends beyond matching aggregate time series. Microeconomic evidence on individual price changes and wage contracts helps assess nominal rigidity; household spending data help assess intertemporal substitution; independently identified policy surprises help evaluate monetary transmission. Forecast comparisons test whether structural restrictions improve predictions relative to simpler statistical alternatives. The empirical status of the synthesis is therefore heterogeneous: evidence for some degree of nominal rigidity does not establish every representative-agent assumption, and success in forecasting ordinary fluctuations does not establish reliability during a financial crisis or a major regime change."
    },
    {
      "heading": "Financial crises, fiscal policy, and the effective lower bound",
      "text": "The global financial crisis exposed major omissions in many benchmark implementations. Basic models often included a risk-free asset but no economically consequential banking sector, default, or intermediary leverage. Such models could represent weaker demand through an adverse shock, yet could not explain how mortgage losses, funding pressures, and impaired intermediary balance sheets generated and amplified that weakness. This was a limitation of prominent baseline specifications rather than a proof that equilibrium analysis could never accommodate financial instability.\n\nFinancial-friction research both preceded and followed the crisis. Ben Bernanke, Mark Gertler, and Simon Gilchrist developed the financial accelerator mechanism, in which weakened borrower net worth raises external financing costs and amplifies reductions in spending. Nobuhiro Kiyotaki and John Moore’s 1997 credit-cycle model emphasized the interaction between collateral values and borrowing capacity. Later New Keynesian models incorporated constrained intermediaries, credit spreads, and disruptions to financial intermediation. These mechanisms make a single policy interest rate an incomplete description of financing conditions.\n\nThe effective lower bound on nominal interest rates introduced another challenge. If the natural real interest rate becomes sufficiently negative, an ordinary policy-rate reduction may be unable to align the expected real rate with it. Expected deflation can make the problem worse by increasing the real cost of borrowing. In such circumstances, the usual stabilizing response to weak demand may be constrained, and adverse feedback between falling spending and falling inflation can arise.\n\nWithin the framework, forward guidance works by changing expected future short-term rates, while quantitative easing requires additional mechanisms if it is to matter beyond signaling. In a frictionless model with fully substitutable assets and unconstrained investors, exchanging one government liability for another need not alter real allocations. Portfolio segmentation, duration risk, borrowing limits, and intermediary constraints can give asset purchases independent effects. The strength of those effects cannot be inferred from the basic three-equation model alone.\n\nThe lower bound also changes the analysis of fiscal policy. Government purchases can support demand more strongly when monetary policy does not raise rates in response, and induced inflation may lower the real interest rate. A fiscal multiplier is nevertheless conditional on the shock’s duration, monetary accommodation, financing, expectations, and household behavior. Fiscal arrangements also affect price-level determination and the feasibility of monetary commitments. The crisis-era extensions therefore weakened any simple division in which monetary policy alone handles stabilization while fiscal and financial institutions remain passive background conditions."
    },
    {
      "heading": "Heterogeneity, information, and the framework’s boundaries",
      "text": "Representative-household models suppress differences in wealth, income risk, borrowing access, and spending responses. A household with substantial liquid savings can smooth consumption after a temporary income loss; another with little liquidity may cut expenditure immediately. Aggregating both into a single optimizing consumer can misrepresent monetary transmission and fiscal effects. Extensions with borrowing-constrained households provided an intermediate step, while heterogeneous-agent New Keynesian models combine nominal rigidity with distributions of household assets and idiosyncratic risks.\n\nIn these models, interest-rate changes affect spending through more than intertemporal substitution. They alter labor income, interest receipts and payments, asset values, and the distribution of resources between households with different propensities to consume. Greg Kaplan, Benjamin Moll, and Giovanni Violante’s 2018 analysis of monetary policy in a heterogeneous-agent setting highlighted the importance of indirect effects operating through changes in labor demand and household income. The distribution of liquidity can consequently influence aggregate policy transmission, not merely the distributional consequences of an otherwise unchanged aggregate response.\n\nInformation assumptions are another boundary. Full-information rational expectations can imply large effects from announcements about policies far in the future, sometimes described as the forward-guidance puzzle. Models incorporating incomplete information, learning, limited attention, or bounded rationality can attenuate those responses. These alternatives require explicit accounts of what agents observe and how they revise beliefs. They also introduce empirical questions about disagreement, forecast errors, and the speed with which households and firms learn about policy changes.\n\nOther extensions reconsider labor-market adjustment, international linkages, and nonlinear behavior. Search-and-matching models distinguish employment relationships from hours worked; open-economy versions examine exchange rates and imported inflation; multisector models allow energy or supply-chain disturbances to affect sectors differently. Large shocks can make local approximations unreliable, particularly when borrowing constraints bind, interest rates reach a lower bound, or price-adjustment frequencies change. Solving such models may require methods that accommodate nonlinearities and occasionally binding constraints.\n\nThe resulting research program has no sharp boundary. Some extensions preserve the original synthesis’s representative-agent core; others retain only its emphasis on explicit intertemporal decisions, nominal adjustment, and equilibrium consistency. Empirical disputes concern the size and persistence of monetary effects, the relevant sources of inflation, the measurement of natural output and interest rates, and the robustness of optimal-policy prescriptions. A particularly difficult unresolved task is distinguishing structural change from a sequence of unusual shocks: either can produce a prolonged departure from the relationships estimated during more stable periods, but they imply different forecasts and policy counterfactuals."
    }
  ],
  "links": [
    "macroeconomics",
    "new-keynesian-economics",
    "dynamic-stochastic-general-equilibrium",
    "is-lm-model",
    "real-business-cycle-theory",
    "microeconomics",
    "rational-expectations",
    "labor-market",
    "monopolistic-competition",
    "marginal-cost",
    "inflation-expectations",
    "phillips-curve",
    "real-interest-rate",
    "taylor-rule",
    "monetary-policy",
    "central-bank",
    "inflation-targeting",
    "central-bank-independence",
    "productivity",
    "gross-domestic-product",
    "european-central-bank",
    "global-financial-crisis",
    "financial-intermediation",
    "deflation",
    "forward-guidance",
    "quantitative-easing",
    "fiscal-policy",
    "fiscal-multiplier",
    "bounded-rationality"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}