## _wp04234 — Section V: Estimated DSGE analysis of technology shocks, nominal and real frictions

## Source details

**Canonical URL:** [_wp04234 — Section V: Estimated DSGE analysis of technology shocks, nominal and real frictions](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04234.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04234.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04234.pdf.json)

---

### H3: Empirical SVAR evidence on technology shocks (baseline)
- Data and sample
  - Quarterly U.S. data, sample period 1948:1–2002:4 (Haver USECON).
  - Output: nonfarm business sector output (LXNFO). Hours: hours of all persons in nonfarm business sector (LXNFH). Population: civilian noninstitutional population aged 16 and over (LNN).
- Baseline SVAR findings (Galí (1999) replication and updates)
  - A positive technology shock (size normalized to one standard deviation) yields:
    - Significant and persistent decline in hours; hours return close to original level only after more than a year.
    - Positive but muted initial response of output.
  - Variance decomposition at business-cycle frequencies (Baxter–King band-pass):
    - Technology shocks account for 5 percent of variance of output and 7 percent of variance of hours.
    - Correlation between output and hours from technology-driven component: -0.08 (versus observed data correlation 0.88).
  - Shocks with no permanent effect on productivity:
    - Account for 95 percent of variance of hours and 93 percent of variance of output at business cycle frequencies.
    - Generate a near-perfect correlation of 0.96 between output and hours.
- Robustness and cross-checks
  - Five-variable and higher-dimensional VARs produce estimates similar to two-variable baseline.
  - Monte Carlo simulation with 500 draws used for standard error bands.
  - Correlation of VAR permanent shock with independent BFK technology measure (annual 1950–89, "fully corrected"): 0.45 (significant at the 5 percent level).
  - OLS regression: BFKt = 0.29 BFKt−1 + 0.67 εzt − 0.32 εdt with t-statistics (1.85), (2.16), (1.11) respectively — supports alignment of VAR permanent shocks with independent technology measure.
  - No significant correlation between VAR permanent shock ztε and capital tax proxies: ΔτJ correlation -0.06; ΔτM correlation 0.12; OLS p-values 0.54, 0.21, 0.68, 0.34 for various lag specifications.

### H3: Sensitivity to labor input transformations and I-shocks vs N-shocks
- Labor transformations (selected outcomes)
  - First-difference per capita hours: typically negative short-run hours response to positive technology shock.
  - Levels of per capita hours: small positive point estimate for impact response of hours, not statistically different from zero in one reported case; technology shocks can account for up to 37 percent of output variance and 11 percent of hours variance in that specification.
  - Quadratic detrending: similar to differences — decline in hours and small contribution of technology shocks (example: 7 percent of variance).
  - Total hours in some specifications: strong and statistically significant negative impact on hours across first differences, levels, and quadratic detrending.
  - Employment and real GDP specifications: uniform short-run decline in employment after a positive technology shock and limited variance contribution to GDP and employment.
- Investment-specific (I-shocks) versus sector-neutral (N-shocks) (Fisher (2003) framework — selected entries)
  - Per capita hours, Difference specification:
    - N-Shocks Var(y) 0.06; Var(n) 0.06; Corr(y,n) -0.09.
    - I-Shocks Var(y) 0.22; Var(n) 0.19; Corr(y,n) 0.94.
  - Per capita hours, Level specification:
    - N-Shocks Var(y) 0.12; Var(n) 0.02; Corr(y,n) 0.16.
    - I-Shocks Var(y) 0.62; Var(n) 0.60; Corr(y,n) 0.96.
  - Conclusion: I-shocks can generate high positive correlation between output and hours; quantitative importance of I-shocks depends on labor input transformation.

### H3: Theoretical channels for negative short-run output–hours comovement
- Nominal-friction channels (monetary policy interaction)
  - Sticky prices (Calvo) combined with typical monetary policy (Taylor-type rule with interest-rate smoothing) can generate short-run decline in hours after a positive technology shock when monetary policy does not fully accommodate productivity increases.
  - Effective policy accommodation index Θ with 0<Θ≤1; output response yψΘ with Θ≤1. Example calibration:
    - κ = 0.024; σ upper bound 1; πφ = 1.5; yφ = 0.5; aρ = 0.95; β = 0.99; α = 1/3.
    - Θ = 0.28; Campbell (1994) benchmark yψ = 1.45 implies nψ = -0.87 (negative employment elasticity).
  - Investment-specific shocks in sticky-price environment tend to raise both output and hours on impact.
- Real-friction channels (no nominal rigidities required)
  - Habit formation combined with capital adjustment costs can produce sluggish consumption/investment, implying output increases less than productivity and a dominant income effect that reduces hours (Francis and Ramey (2003a)).
  - No substitutability between labor and capital can produce short-run decline in hours.
  - Slow diffusion of technology (Rotemberg (2003)) or open-economy terms-of-trade effects (Collard and Dellas (2002)) can generate similar short-run declines in hours.
- Micro evidence
  - Firm-level panel (Marchetti and Nucci (2004)): negative employment response to firm-level TFP shocks is larger for “sticky” price firms than for “flexible” price firms — consistent with nominal-friction channel.

### H3: Estimated DSGE model — setup, priors, and posterior estimates
- Model features and observables
  - Habit formation, Calvo staggered price and wage setting, indexation to lagged inflation, Taylor-type interest-rate rule with smoothing, unit root in technology, exogenous shocks: technology, preference, price markup, wage markup, monetary shock.
  - Observables: output growth, inflation, nominal interest rate, hours, real wage-output ratio.
  - Sample: 1948:1–2002:4. Wages: compensation per hour in nonfarm business sector (LXNFC). Price deflator: nonfarm business sector deflator (LXNFI). Interest rate: three-month Treasury bill (FTB3).
  - Hours and real wage-output ratio detrended with quadratic trend; inflation, output growth, nominal interest rate treated as stationary deviations from sample mean.
- Bayesian estimation details
  - Priors: uniform for many structural parameters; gamma for shock standard deviations; AR(1) coefficients uniform between 0 and 0.97.
  - Fixed parameters: β = 0.99; product and labor demand elasticities set to 6 (implying steady-state markup 20 percent).
  - Posterior sampling: Metropolis-Hastings, 500,000 draws; Kalman filter for likelihood evaluation.
- Selected posterior means (with interpretation)
  - Habit formation parameter b: 0.42 (posterior mean).
  - Elasticity of marginal disutility of hours φ: 0.80.
  - Price stickiness (θ_p): posterior mean 0.53 (implies average price contract duration slightly above two quarters).
  - Wage stickiness (θ_w): posterior mean 0.05 (very low).
  - Price indexation η_p: 0.04 (low). Wage indexation η_w: 0.42.
  - Interest-rate smoothing ρ_r: 0.69. Response to inflation φ_π: 1.33. Small response to output growth φ_y: 0.26.
  - Shock AR(1) persistence: high (ρ_g 0.93; ρ_u 0.95; ρ_v 0.91; price-markup ρ between 0.95 and 0.91 as noted).

### H3: Model fit, impulse responses, and variance decompositions
- Selected second-moment comparisons (original data / model / technology component / contribution in percent)
  - Output growth: Data 1.36; Model 1.27; Technology component 0.60; Contribution 22.3
  - Inflation: Data 0.72; Model 0.73; Technology component 0.18; Contribution 6.0
  - Interest rate: Data 0.72; Model 0.67; Technology component 0.04; Contribution 0.3
  - Hours: Data 3.11 percent; Model 4.60 percent; Technology component 0.42; Contribution 0.8
  - Real wage-output: Data 3.69 percent; Model 4.44 percent; Technology component 0.13; Contribution 0.1
  - Correlation between Δy and Δn: Data 0.75; Model 0.72; Technology component -0.49
- Band-pass filtered (business-cycle) moments
  - Output (BP): Data 2.04; Model 2.04; Technology component 0.87; Contribution 18.2
  - Hours (BP): Data 1.69; Model 1.69; Technology component 0.26; Contribution 2.3
  - Correlation (y,n) BP: Data 0.88; Model 0.88; Technology component -0.14
- Impulse responses to a one-standard-deviation permanent technology shock (posterior mean)
  - Hours: drop on impact by about 0.4 percentage points; converge monotonically back to initial level thereafter (about four quarters for output to reach new steady-state).
  - Output: gradual adjustment to permanently higher plateau; about four quarters to reach new steady-state.
- Full variance decomposition (contributions in percent, selected)
  - Output growth: Monetary 4.8; Technology 22.3; Preference 57.1; Price markup 8.0; Wage markup 7.1
  - Inflation: Monetary 27.1; Technology 6.1; Preference 36.3; Price markup 13.7; Wage markup 14.7
  - Nominal rate: Monetary 5.0; Technology 0.4; Preference 72.3; Price markup 9.8; Wage markup 11.8
  - Hours: Monetary 0.4; Technology 0.8; Preference 70.0; Price markup 17.6; Wage markup 9.6
  - Real wage-output: Monetary 0.1; Technology 0.1; Preference 73.6; Price markup 12.0; Wage markup 12.8
- Correlations implied by alternative model specifications (BP-filtered)
  - Original: -0.14
  - Flexible wages: -0.16
  - Flexible prices: -0.18
  - No habit formation: -0.29
  - Flexible prices and wages: -0.21
  - No frictions (RBC): 0.22
  - Inflation targeting: -0.15
- Structural counterfactuals (shut-down experiments)
  - Removing only nominal or only real frictions does not reproduce positive RBC-like correlation; only when all rigidities are shut down does the model produce correlation 0.22 (consistent with pure RBC).
  - Interpretation: interaction of nominal rigidities (predominantly sticky prices) and real rigidities (habit formation and related features) is required to account for negative comovement between hours and output in the estimated model.
  - Preference/demand shocks dominate postwar business-cycle variability in the model.

### H3: Conclusions, policy-relevant implications, and guidance for research
- Empirical synthesis
  - SVARs, growth-accounting, industry-level studies, and the estimated DSGE model provide little support for sector-neutral technology shocks as the quantitatively dominant source of postwar U.S. business cycles.
  - Sector-neutral technology shocks typically explain only a small fraction of business-cycle variance of output and hours and often induce a decline in labor input on impact in many empirical specifications.
  - Investment-specific technology shocks (I-shocks) can generate the positive output–hours comovement characteristic of business cycles; their quantitative importance is sensitive to labor-input measurement choices.
- Policy implications
  - The response of employment to technology shocks depends critically on monetary policy reaction functions and the presence of nominal rigidities (effective accommodation index Θ).
  - Monetary policy that does not sufficiently accommodate permanent productivity gains can lead to output increases smaller than productivity gains and short-run declines in hours.
  - Habit formation and other real rigidities can produce short-run consumption/leisure trade-offs leading to declines in hours after productivity improvements.
  - Given the estimated dominance of preference/demand shocks, stabilization-focused policy (monetary and fiscal) is crucial to manage business-cycle fluctuations.
- Research and modeling guidance
  - SVAR evidence robust across identification schemes should inform DSGE model development; models must match conditional impulse responses, not only unconditional second moments.
  - Careful treatment of measurement and stochastic properties of labor series is essential: overdifferencing or incorrect detrending can distort inferences about technology shocks.
  - The Solow residual is a noisy, procyclical proxy for technology; independent measures (e.g., BFK "fully corrected" series) and structural identification help validate technological interpretation of long-run VAR shocks.

*Source: _wp04234 — Section V lays out and analyzes an estimated dynamic stochastic general equilibrium (DSGE) model that incorporates both nominal and real frictions, and evaluates their respective role (PDF: _wp04234).*

### References..............................................................................................................

### _wp04234 - References

### I. Introduction and framing
- Seminal works: Kydland and Prescott (1982) and Prescott (1986a) established the RBC paradigm attributing a central role to exogenous technology variations as a source of economic fluctuations.
- Core RBC claim: calibrated neoclassical growth models with consumption-leisure choice and stochastic changes in total factor productivity can account for the bulk of postwar U.S. fluctuations by matching unconditional second moments (relative standard deviations and correlations).
- Notable quotation preserved from the source: Prescott (1986b) claimed “...that technology shocks account for more than half the fluctuations in the postwar period, with a best point estimate near 75 percent.”
- Evolution of the field: while RBC modeling tools have been broadly adopted, two substantive critiques emerged:
  - The profession emphasizes the importance of monetary policy and nominal frictions for short-run macro phenomena.
  - The centrality of technological change as the dominant cyclical force has been questioned; this paper focuses on literature challenging that view.

### II. Empirical focus and identification strategies
- Research agenda emphasized: obtaining evidence on the role of technology that is “more direct” than model-based moment-matching.
- Two principal empirical approaches discussed:
  - Structural VAR identification exploiting the permanent nature of technology-driven productivity changes.
  - Use of more direct measures of technological change and examination of their comovements with macro variables.
- Data transformations and filtering explicitly cited:
  - Business-cycle fluctuations interpreted using the Baxter and King (1999) band-pass filter calibrated to remove fluctuations of periodicity outside an interval between 6 and 32 quarters.
  - Time series examples use the U.S. nonfarm business sector series for hours and output over the period 1948:1-2002:4 (also written as sample period 1948:01–2002:04).

### III. Key empirical phenomena and model performance
- Robust descriptive fact motivating the literature: a strong positive comovement between output and labor input measures in industrial economies.
- RBC model performance:
  - The basic RBC model can generate fluctuations in labor input and output of magnitude, persistence, and degree of comovement roughly similar to the BP-filtered series.
  - When the actual sequence of technology shocks (proxied by estimated disturbances of an AR process for the Solow residual) is used as input, model-implied equilibrium paths of output and labor input track observed historical patterns closely — a stricter test than moment-matching.

### IV. Central research questions posed by the reviewed literature
- What have been the effects of technology shocks in the postwar U.S. economy?
- How do these effects differ from standard RBC model predictions?
- What is the contribution of technology shocks to business cycle fluctuations?
- What model features are required to account for observed effects?

### V. Overall assessment from the reviewed literature
- The literature presents tentative and sometimes contradictory answers.
- The bulk of the evidence reviewed provides little support for the initial RBC claim of a central role for technological change as the primary source of business cycles in the postwar U.S. economy.

### VI. Structure of the rest of the paper (as described)
- Section II: reviews early papers questioning the importance of technology shocks and presents basic evidence regarding their effects.
- Section III: discusses criticisms and possible pitfalls of that literature.
- Section IV: presents the case for nominal frictions as an explanation of estimated effects of technology shocks, and summarizes real explanations for the same effects found in the literature.

*Source: _wp04234 - References*

### Section V lays out and analyzes an estimated dynamic stochastic general equilibrium

### _wp04234 - Section V lays out and analyzes an estimated dynamic stochastic general equilibrium (DSGE) model that incorporates both nominal and real frictions, and evaluates their respective role

### Estimated effects of technology shocks (SVAR evidence)
- Data and baseline
  - Quarterly U.S. data for the period 1948:1-2002:4 from Haver USECON.
  - Output: nonfarm business sector output (LXNFO). Hours: hours of all persons in nonfarm business sector (LXNFH). Population: civilian noninstitutional population aged 16 and over (LNN).
  - Baseline VAR specification: first-differences for hours (tt nnˆ∆ =), VAR for [tt nx∆∆ , ] (difference specification).
- Main empirical findings (Galí (1999) replication and updates)
  - A positive technology shock (size normalized to one standard deviation) yields:
    - A significant and persistent decline in hours; point estimates imply hours return close to original level only after more than a year.
    - A positive but muted initial response of output.
  - Variance decomposition at business cycle frequencies (band-pass filtered):
    - Technology shocks account for 5 percent of variance of output and 7 percent of variance of hours.
    - Correlation between output and hours from technology-driven component: -0.08 (essentially zero), versus observed data correlation 0.88.
  - Shocks with no permanent effect on productivity:
    - Account for 95 percent of variance of hours and 93 percent of variance of output at business cycle frequencies.
    - Generate a near-perfect correlation of 0.96 between output and hours.
- Robustness and extensions
  - Five-variable VARs and higher-dimensional specifications produce estimates similar to the baseline two-variable VAR.
  - Monte Carlo simulation used to obtain standard error bands based on 500 draws from reduced-form VAR distribution.

### Related empirical evidence and cross-checks
- Independent methodologies generally support the SVAR evidence:
  - Basu, Fernald, and Kimball (1999) (BFK): sophisticated growth-accounting measure of aggregate technological change shows sharp short-run decline in inputs (including labor) when technology increases, with output showing no significant immediate change.
  - Kiley (1997), Francis (2001), Franco and Philippon (2004): industry- and sector-level analyses typically find negative or non-positive short-run employment responses to positive industry-specific technology shocks and limited aggregate correlation across industries.
  - Shea (1998): industry innovations rarely cause significant TFP increases; where TFP increases are detected inputs respond opposite to TFP.
  - Cross-country evidence: negative employment response to positive technology shocks in G-7 (except Japan); euro-area estimates show technology shocks account for 5 percent and 9 percent of variance of employment and output, respectively, with correlation -0.67.
- Tests against capital tax mislabeling
  - Correlation coefficients between VAR permanent shock ztε and first-difference capital tax proxies:
    - Correlation with Jones series ΔτJ: -0.06 (not significant).
    - Correlation with McGrattan series ΔτM: 0.12 (not significant).
  - OLS regressions of tax changes on current and lagged VAR permanent shocks yield p-values 0.54 (four lags, Jones), 0.21 (eight lags, Jones), 0.68 (four lags, McGrattan), 0.34 (eight lags, McGrattan) — no evidence that permanent VAR shocks capture capital tax innovations.
- Correlation with BFK technology measure (annual 1950–89, "fully corrected" series)
  - Correlation between VAR permanent shock and BFK: 0.45 (significant at the 5 percent level).
  - Correlation between VAR transitory shock and BFK: -0.04 (insignificant).
  - OLS regression (normalized BFK on its own lag and contemporaneous permanent and transitory shocks):
    - Estimated: BFKt = 0.29 BFKt−1 + 0.67 εzt − 0.32 εdt
    - t-statistics in brackets: (1.85) for lag, (2.16) for εz, (1.11) for εd — reinforces that permanent VAR shocks align with independent technological-change measures.

### Robustness to alternative labor input transformations
- Specification sensitivity (per capita hours and total hours; levels, first differences, quadratic detrending)
  - First-difference specification typically yields negative short-run hours response to positive technology shock.
  - Level specification of per capita hours yields a small positive point estimate for impact response of hours, but the response is not statistically different from zero; technology shocks still account for relatively small fractions of business-cycle variance (examples: under level specification technology shocks account for 37 percent of output variance and 11 percent of hours variance at business cycle frequencies in one reported case).
  - Quadratic detrending of per capita hours yields results similar to difference specification: decline in hours and small contribution of technology shocks (e.g., 7 percent of variance).
  - Using total hours (non-normalized) in first differences, levels, and quadratic detrending produces a strong and statistically significant negative impact on hours for all three transformations in some specifications.
- Employment and GDP specifications
  - Using employment as labor input and real GDP as output yields uniform results: decline in employment in the short run following a positive technology shock across transformations, and limited contribution of technology shocks to variance of GDP and employment.
- Investment-specific vs. sector-neutral technology shocks
  - Fisher (2003) framework distinguishing N-shocks (sector-neutral) and I-shocks (investment-specific):
    - N-shocks: negligible contribution to variance of output and hours, low correlation between output and hours.
    - I-shocks: generate high positive correlation between output and hours (consistent with business-cycle comovement). Contribution of I-shocks to variance depends on labor input transformation: in levels it can account for more than half of variance of output and hours; in differences or detrended specifications contribution falls below one-fourth.

### Explanations for the "anomalous" negative comovement (theory)
- Broad classification: nominal-friction channels and real-friction channels.
- Nominal frictions + monetary policy
  - Sticky prices (Calvo pricing) combined with monetary policy rules (e.g., Taylor-type rules) can generate a short-run decline in hours after a positive technology shock, unless monetary policy fully accommodates the productivity increase.
  - Simple illustrative New Keynesian model (Calvo staggering, forward-looking IS, Phillips curve, Taylor rule with interest rate smoothing):
    - Natural output and employment respond positively to productivity (yψ, nψ from an underlying RBC-like real model).
    - Effective policy accommodation index Θ derived; 0<Θ≤1 under weak assumption aρφπ > (paper uses condition aρφπ >; full text: aρφπ >).
    - Output response to technology shock is yψΘ with Θ≤1; employment response bounded similarly and can be negative depending on parameters.
  - Calibration example:
    - Rotemberg and Woodford estimate: κ = 0.024.
    - Upper bound for σ often set to 1.
    - Taylor-rule coefficients πφ = 1.5 and yφ = 0.5 (Taylor's values).
    - Technology persistence aρ = 0.95; β = 0.99; α = 1/3.
    - With these, Θ = 0.28 and using Campbell (1994) benchmark yψ = 1.45 yields implied nψ = -0.87 (negative employment elasticity).
  - Investment-specific technology shocks (I-shocks) in a sticky-price environment tend to raise both output and hours on impact because present efficiency of labor is unchanged and firms must employ inputs to produce current output.
- Real-friction explanations (no nominal rigidities required)
  - Habit formation and capital adjustment costs (Francis and Ramey (2003a)):
    - Introduce sluggish consumption/investment response; output may increase less than productivity and hours can decline via a dominant income effect (households consume more leisure).
  - No substitutability between labor and capital (Francis and Ramey variant): short-run decline in hours when less labor per workweek is needed despite higher output opportunities.
  - Slow diffusion of technology (Rotemberg (2003)): gradual adoption generates wealth effects that increase leisure, producing short-run declines in hours and output.
  - Open-economy terms-of-trade effects (Collard and Dellas (2002)): positive domestic technology shock can deteriorate terms of trade and prompt leisure increases if domestic-foreign substitution is low.
- Micro evidence consistent with nominal-friction channel
  - Marchetti and Nucci (2004): firm-level panel of Italian manufacturing — negative employment response to firm-level TFP shocks is larger for “sticky” price firms than for “flexible” price firms.

### Estimated DSGE model with both nominal and real frictions (Bayesian estimation)
- Model features
  - Habit formation in consumption, Calvo staggered price and wage setting, indexation to lagged inflation, Taylor-type interest rate rule with interest rate smoothing, unit root in technology process, exogenous shocks: technology, preference, price markup, wage markup, monetary shock.
  - Observables: output growth, inflation, nominal interest rate, hours, real wage-output ratio.
  - Sample: 1948:1-2002:4; series for wages: compensation per hour in nonfarm business sector (LXNFC); price deflator: nonfarm business sector deflator (LXNFI); interest rate: three-month Treasury bill (FTB3).
  - Hours and real wage-output ratio detrended with quadratic trend; inflation, output growth, nominal interest rate treated as stationary deviations from sample mean.
- Bayesian estimation approach
  - Priors: uniform for many structural parameters (habit, Calvo probabilities, indexation), gamma for shock standard deviations, AR(1) coefficients uniform between 0 and 0.97. Fixed parameters: β = 0.99; product and labor demand elasticities set to 6 (implying steady-state markup 20 percent).
  - Posterior sampling: Metropolis-Hastings, 500,000 draws; use Kalman filter for likelihood evaluation.
- Posterior parameter estimates (selected)
  - Habit formation parameter: 0.42 (posterior mean).
  - Elasticity of marginal disutility of hours φ: 0.80.
  - Price contract average duration: slightly above two quarters (implied by price stickiness parameter).
  - Wage stickiness estimate: very low (Calvo wage parameter small).
  - Price indexation: 0.04 (low). Wage indexation: 0.42.
  - Interest-rate smoothing: 0.69. Response to inflation: 1.33 (lean-against-the-wind). Small response to output growth in rule.
  - Shock AR(1) coefficients: high persistence (between 0.95 for price markup shock and 0.91 for wage markup shock).
- Model fit (selected moments)
  - Model replicates standard deviations of output, inflation, and nominal interest rate well.
  - Unconditional correlation between growth rates of hours and output:
    - Data: 0.75.
    - Model: 0.72.
  - Model overestimates standard deviation of hours:
    - Data: 3.11 percent.
    - Model: 4.6 percent.
  - Real wage-to-output ratio standard deviation:
    - Data: 3.69 percent.
    - Model: 4.44 percent.

### DSGE model implications for technology shocks and business-cycle contributions
- Impulse responses to a one-standard-deviation permanent technology shock (posterior mean)
  - Hours: drop on impact by about 0.4 percentage points; converge monotonically back to initial level thereafter (about four quarters for output to reach new steady-state).
  - Output: gradual adjustment to permanently higher plateau; about four quarters to reach new steady-state.
- Variance contributions (technology shocks only)
  - Technology shocks explain:
    - 22 percent of variability of output growth.
    - 6 percent of variability of inflation.
    - Insignificant amount of overall volatility in hours and some other variables.
  - Correlation between Δy and Δn when only technology shocks drive the model: -0.49 (negative), contrasting with actual filtered series correlation of 0.88.
  - Band-pass filtered business-cycle statistics (model with technology shocks only):
    - Technology shocks explain only a small fraction of the variance of business-cycle components of output and hours.
    - Conditional correlation between output and hours from model's technology-driven components: -0.14 (versus 0.88 observed).
- Full variance decomposition (Table 6 summary)
  - Preference (demand) shock:
    - Explains above 70 percent of variance of hours, real wage-output ratio, and nominal interest rate.
    - Explains 57 percent of variance of output and 36 percent of variance of inflation.
  - Monetary shock:
    - Explains approximately 5 percent of output growth and nominal interest rate variance.
    - Explains 27 percent of inflation variability.
  - Price and wage markup shocks:
    - Contribute between 7 percent and 17 percent of variance across variables.
  - Technology shock:
    - Explains about 22 percent of output growth variance but small parts of other variables’ variance (including hours).
  - Overall: preference/demand shocks are the dominant source of postwar business-cycle variability in the model.

### Structural counterfactuals: which frictions generate the negative output–hours comovement?
- Counterfactual experiments (shut down features and simulate with same estimated remaining parameters) — correlation between business-cycle components of output and hours:
  - Shutting down sticky wages (θw = ηw = 0) — correlations remain large and negative.
  - Shutting down sticky prices but maintaining sticky wages and habits — correlations remain large and negative.
  - Flexible prices and wages with habit formation — similar negative correlation persists.
  - Shutting down habit formation but keeping nominal rigidities — conditional and unconditional correlations still mimic data signs.
  - Only when all rigidities (nominal and real) are shut down does the model produce the positive correlation consistent with a pure RBC model.
  - Removing the output response in the Taylor rule (central bank responds only to inflation) does not eliminate negative conditional correlation.
- Interpretation
  - Both real rigidities (habit formation) and nominal rigidities (predominantly sticky prices) are relevant to account for the negative comovement between hours and output following technology shocks in the estimated model.
  - Neither nominal nor real frictions alone fully explain the empirical pattern; the interaction matters.
  - Preference/demand shocks remain primary drivers of business-cycle comovement between output and labor input in the estimated model.

### Conclusions and policy-relevant implications
- Empirical synthesis
  - Bulk of evidence (SVARs, growth-accounting, industry-level studies, estimated DSGE) casts serious doubt on aggregate, sector-neutral technology shocks as a quantitatively dominant source of postwar U.S. business cycles.
  - Technology shocks (sector-neutral) typically explain only a small fraction of business-cycle variance of output and hours and tend to induce a decline in labor input on impact in many empirical specifications.
  - Investment-specific technology shocks (I-shocks) can produce the positive output–hours comovement characteristic of business cycles and may play a more important role if labor input is measured in levels; however their quantitative importance depends on labor transformation choices.
- Mechanisms and policy implications
  - The response of employment to technology shocks is not invariant: it depends critically on monetary policy reaction functions and the presence of nominal rigidities.
  - Monetary policy that does not sufficiently accommodate permanent productivity gains (limited effective accommodation Θ) can lead to output increases that are smaller than productivity gains and to short-run declines in hours.
  - Habit formation and other real rigidities can generate short-run consumption/leisure trade-offs that produce declines in hours following productivity improvements.
  - Given the dominance of preference/demand shocks in explaining cyclical volatility in the estimated DSGE, stabilization-focused policy (monetary and fiscal) remains crucial to manage business-cycle fluctuations.
- Research and model-building guidance
  - SVAR evidence that is robust across identification schemes should inform DSGE model development; models must match conditional impulse responses, not only unconditional second moments.
  - Careful treatment of measurement and of the order of integration of labor series is essential: mis-specifying the stochastic properties of hours (overdifferencing or assuming stationarity when the data display unit roots or quadratic trends) can distort inferences about technology shocks.
  - The Solow residual is a noisy, procyclical proxy for technology; independent measures (e.g., BFK "fully corrected" series) and structural identification help validate the technological interpretation of long-run VAR shocks.

*Italic source: _wp04234 - Section V lays out and analyzes an estimated dynamic stochastic general equilibrium (DSGE) model that incorporates both nominal and real frictions, and evaluates their respective role (PDF: _wp04234 - Section V lays out and analyzes an estimated dynamic stochastic general equilibrium).*

### Section V of the present paper, following the footsteps of a number of authors referred to in

### Section V of the present paper, following the footsteps of a number of authors referred to in

### Methodological note and response to McGrattan
- The authors implemented a model (referred to by McGrattan as the “triple-sticky” model) as an illustration of an approach to analyze how different frictions shape estimated effects of technology shocks.
- The authors’ objective was illustrative rather than to build a full-fledged economy model encompassing all aspects such as capital accumulation.
- Other literature uses richer structures that include endogenous capital accumulation and nest the standard RBC model; some of those papers (e.g., Smets and Wouters, 2003b) have analyzed technology-shock effects and find a negative response of hours to a positive technology shock.
- The authors conjecture that McGrattan’s “triple-sticky model with investment” would imply a similar negative hours response, though McGrattan reports no comparable evidence.

### Empirical findings — Effects of technology shocks on output and hours (Nonfarm Business Sector)
- Table 1: Contribution to Var(y), Var(n), Corr(y,n) and Conditional Impact on n and y (Sign and Significance)
  - Per Capita Hours
    - Difference: Var(y) 0.07; Var(n) 0.05; Corr(y,n) -0.08; Sign -  / +; Significance yes / yes
    - Level: Var(y) 0.37; Var(n) 0.11; Corr(y,n) 0.80; Sign + / +; Significance no / yes
    - Detrended: Var(y) 0.07; Var(n) 0.05; Corr(y,n) -0.11; Sign - / +; Significance yes / yes
  - Total Hours
    - Difference: Var(y) 0.06; Var(n) 0.06; Corr(y,n) -0.03; Sign - / +; Significance yes / yes
    - Level: Var(y) 0.10; Var(n) 0.36; Corr(y,n) 0.80; Sign - / -; Significance yes / no
    - Detrended: Var(y) 0.15; Var(n) 0.36; Corr(y,n) 0.80; Sign - / 0; Significance yes / no

### Empirical findings — Effects of technology shocks on GDP and employment
- Table 2: Contribution to Var(y), Var(n), Corr(y,n) and Conditional Impact on n and y (Sign and Significance)
  - Employment Rate
    - Difference: Var(y) 0.31; Var(n) 0.04; Corr(y,n) 0.40; Sign -  / +; Significance yes / yes
    - Level: Var(y) 0.03; Var(n) 0.19; Corr(y,n) -0.30; Sign - / +; Significance yes / no
    - Detrended: Var(y) 0.15; Var(n) 0.04; Corr(y,n) -0.43; Sign - / +; Significance yes / yes
  - Total Employment
    - Difference: Var(y) 0.21; Var(n) 0.03; Corr(y,n) -0.40; Sign - / +; Significance yes / yes
    - Level: Var(y) 0.09; Var(n) 0.08; Corr(y,n) -0.72; Sign - / +; Significance yes / yes
    - Detrended: Var(y) 0.09; Var(n) 0.09; Corr(y,n) -0.68; Sign - / +; Significance yes / no

### Investment-specific technology shocks — The Fisher model (Table 3)
- Per Capita Hours (N-Shocks vs I-Shocks)
  - Difference: N-Shocks Var(y) 0.06; Var(n) 0.06; Corr(y,n) -0.09. I-Shocks Var(y) 0.22; Var(n) 0.19; Corr(y,n) 0.94.
  - Level: N-Shocks Var(y) 0.12; Var(n) 0.02; Corr(y,n) 0.16. I-Shocks Var(y) 0.62; Var(n) 0.60; Corr(y,n) 0.96.
  - Detrended: N-Shocks Var(y) 0.08; Var(n) 0.07; Corr(y,n) -0.03. I-Shocks Var(y) 0.10; Var(n) 0.09; Corr(y,n) 0.94.
- Total Hours
  - Difference: N-Shocks Var(y) 0.07; Var(n) 0.06; Corr(y,n) 0.05. I-Shocks Var(y) 0.16; Var(n) 0.14; Corr(y,n) 0.94.
  - Level: N-Shocks Var(y) 0.05; Var(n) 0.15; Corr(y,n) 0.33. I-Shocks Var(y) 0.82; Var(n) 0.78; Corr(y,n) 0.97.
  - Detrended: N-Shocks Var(y) 0.10; Var(n) 0.28; Corr(y,n) 0.62. I-Shocks Var(y) 0.09; Var(n) 0.08; Corr(y,n) 0.93.
- Employment Rate
  - Difference: N-Shocks Var(y) 0.21; Var(n) 0.05; Corr(y,n) 0.08. I-Shocks Var(y) 0.19; Var(n) 0.13; Corr(y,n) 0.93.
  - Level: N-Shocks Var(y) 0.08; Var(n) 0.08; Corr(y,n) -0.32. I-Shocks Var(y) 0.86; Var(n) 0.89; Corr(y,n) 0.95.
  - Detrended: N-Shocks Var(y) 0.06; Var(n) 0.17; Corr(y,n) -0.11. I-Shocks Var(y) 0.12; Var(n) 0.10; Corr(y,n) 0.92.
- Total Employment
  - Difference: N-Shocks Var(y) 0.19; Var(n) 0.06; Corr(y,n) -0.05. I-Shocks Var(y) 0.10; Var(n) 0.06; Corr(y,n) 0.90.
  - Level: N-Shocks Var(y) 0.04; Var(n) 0.16; Corr(y,n) -0.25. I-Shocks Var(y) 0.64; Var(n) 0.52; Corr(y,n) 0.96.
  - Detrended: N-Shocks Var(y) 0.04; Var(n) 0.20; Corr(y,n) 0.05. I-Shocks Var(y) 0.12; Var(n) 0.09; Corr(y,n) 0.90.

### Prior and posterior distributions (Table 4) — selected parameter summaries
- b (Uniform(0,1)): Prior mean 0.50 s.d. 0.289; Posterior mean 0.42 s.d. 0.04
- φ (Normal(1,0.25)): Prior mean 1.00 s.d. 0.25; Posterior mean 0.80 s.d. 0.11
- θ_p (Uniform(0,0.9)): Prior mean 0.45 s.d. 0.259; Posterior mean 0.53 s.d. 0.03
- θ_w (Uniform(0,0.9)): Prior mean 0.45 s.d. 0.259; Posterior mean 0.05 s.d. 0.02
- η_p (Uniform(0,1)): Prior mean 0.50 s.d. 0.289; Posterior mean 0.02 s.d. 0.02
- η_w (Uniform(0,1)): Prior mean 0.50 s.d. 0.289; Posterior mean 0.42 s.d. 0.28
- ρ_r (Uniform(0,0.97)): Prior mean 0.485 s.d. 0.284; Posterior mean 0.69 s.d. 0.04
- φ_y (Normal(0.5,.125)): Prior mean 0.50 s.d. 0.13; Posterior mean 0.26 s.d. 0.06
- φ_π (Normal(1.5,0.25)): Prior mean 1.50 s.d. 0.25; Posterior mean 1.35 s.d. 0.13
- ρ_g (Uniform(0,0.97)): Prior mean 0.485 s.d. 0.284; Posterior mean 0.93 s.d. 0.02
- ρ_u (Uniform(0,0.97)): Prior mean 0.485 s.d. 0.284; Posterior mean 0.95 s.d. 0.02
- ρ_v (Uniform(0,0.97)): Prior mean 0.485 s.d. 0.284; Posterior mean 0.91 s.d. 0.01
- σ_z (Gamma(25,0.0001)): Prior mean 0.0025 s.d. 0.0005; Posterior mean 0.003 s.d. 0.000
- σ_a (Gamma(25,0.0004)): Prior mean 0.01 s.d. 0.002; Posterior mean 0.009 s.d. 0.001
- σ_g (Gamma(16,0.00125)): Prior mean 0.02 s.d. 0.005; Posterior mean 0.025 s.d. 0.0024
- σ_u (Gamma(4,0.0025)): Prior mean 0.01 s.d. 0.005; Posterior mean 0.011 s.d. 0.001
- σ_v (Gamma(4,0.0025)): Prior mean 0.01 s.d. 0.005; Posterior mean 0.012 s.d. 0.001

### Second moments and model fit (Table 5)
- Original Data / Model / Technology Component / Contribution to Variance of Each Variable by Technology Shocks
  - Output growth: Data 1.36; Model 1.27; Technology component 0.60; Contribution 22.3
  - Inflation: Data 0.72; Model 0.73; Technology component 0.18; Contribution 6.0
  - Interest rate: Data 0.72; Model 0.67; Technology component 0.04; Contribution 0.3
  - Hours: Data 3.11; Model 4.60; Technology component 0.42; Contribution 0.8
  - Real wage-output: Data 3.69; Model 4.44; Technology component 0.13; Contribution 0.1
  - Correlation between (dy,dn): Data 0.75; Model 0.72; Technology component -0.49
- Band-Pass Filtered Data
  - Output: Data 2.04; Model 2.04; Technology component 0.87; Contribution 18.2
  - Hours: Data 1.69; Model 1.69; Technology component 0.26; Contribution 2.3
  - Correlation between (y,n): Data 0.88; Model 0.88; Technology component -0.14

### Variance decomposition from the estimated DSGE model (Table 6) — contributions in percent
- Output growth: Monetary 4.8; Technology 22.3; Preference 57.1; Price markup 8.0; Wage markup 7.1
- Inflation: Monetary 27.1; Technology 6.1; Preference 36.3; Price markup 13.7; Wage markup 14.7
- Nominal rate: Monetary 5.0; Technology 0.4; Preference 72.3; Price markup 9.8; Wage markup 11.8
- Hours: Monetary 0.4; Technology 0.8; Preference 70.0; Price markup 17.6; Wage markup 9.6
- Real wage-output: Monetary 0.1; Technology 0.1; Preference 73.6; Price markup 12.0; Wage markup 12.8

### Correlations implied by alternative model specifications (Table 7, BP-filtered data)
- Original: -0.14
- Flexible wages: -0.16
- Flexible prices: -0.18
- No habit formation: -0.29
- Flexible prices and wages: -0.21
- No frictions (RBC): 0.22
- Inflation targeting: -0.15

### Graphical and impulse-response evidence (figures referenced)
- Figure 1: Business cycle fluctuations in output and hours (solid line output; dashed line hours).
- Figure 2: Estimated effects of technology shocks (Difference specification, sample period 1948:01–2002:04).
- Figure 3: Sources of business cycle fluctuations (Difference specification, sample period 1948:01–2002:04).
- Figure 5: Technology shocks: VAR versus BFK (solid line VAR technology measure; dashed line BFK technology measure).
- Figure 7: Posterior impulse responses to a technology shock, estimated DSGE model — responses shown for Output, Hours, Interest Rates, Inflation, Output Growth, and Technology over 0–12 quarters after shock; vertical axes labeled in % Dev. from S.S. or % Dev. from S.S. as appropriate.
- Figure 8: Model-based estimates of the role of technology shocks in U.S. postwar fluctuations for Output (Band-Pass) and Hours (Band-Pass); solid line technology component (BP-filtered); dashed line U.S. data (BP-filtered).

*Source: Authors’ calculations.*

### REFERENCES

### _wp04234 - REFERENCES

### Technology shocks and RBC/DSGE literature
- Altig, D., L. J. Christiano, M. Eichenbaum, and J. Linde, 2003, “Monetary Policy and the Dynamic Effects of Technology Shocks” (unpublished, Evanston: Northwestern University).
- Basu, S., J. Fernald, and M. Kimball, 1999, “Are Technology Improvements Contractionary?,” (unpublished, Ann Arbor: University of Michigan).
- Basu, S., and M. Kimball, 1997, “Cyclical Productivity with Unobserved Input Variation,” NBER Working Paper 5915 (Cambridge: National Bureau of Economic Research).
- Burnside, C., and M. Eichenbaum, 1996, “Factor Hoarding and the Propagation of Business Cycle Shocks, American Economic Review, Vol. 86, No. 5, pp. 1154-74.
- Burnside, C., M. Eichenbaum, and S. Rebelo, 1995, “Capital Utilization and Returns to Scale,” NBER Macroeconomics Annual 1995, Ben S. Bernanke and Julio J. Rotemberg,eds.,Cambridge: MIT Press, pp. 67-110.
- Green wood, J., Z. Hercowitz, and P. Krusell, 1997, “Long-Run Implications of Investment Specific Technological Change,” American Economic Review, Vol 87, pp. 342–362.
- Greenwood, J., Z. Hercowitz, and G.W. Huffman, 1998, “Investment, Capacity Utilization, and the Real Business Cycle,” American Economic Review, Vol. 78, No. 3, pp. 402–17.
- Greenwood, J., Z. Hercowitz, and P. Krusell, 2000, “The Role of Investment-Specific Technical Change in the Business Cycle,” European Economic Review, Vol 44, pp. 91–115.
- Kydland, F.E. and E.C. Prescott, 1982, “Time to Build and Aggregate Fluctuations,” Econometrica, Vol. 50, No. 6 ,pp. 1345–70.
- King, R., C. Plosser, and S. Rebelo, 1988a, “Production, Growth and Business Cycles I: The Basic Neoclassical Model,” Journal of Monetary Economics, Vol. 21, March/May, pp. 195-232.
- King, R., C. Plosser, and S. Rebelo, 1988b, “Production, Growth and Business Cycles II: New Directions,” Journal of Monetary Economics, Vol. 21, March/May, pp. 309-341.
- King, R.G., and S.T. Rebelo, 1999, “Resuscitating Real Business Cycles,” in J.B. Taylor and M. Woodford (eds.) Handbook of Macroeconomics, Vol. 1B, pp. 928–1002 (also National Bureau of Economic Research Working Paper 7534).
- Prescott, E.C., 1986a, “Theory Ahead of Business Cycle Measurement,” Quarterly Review 10, pp. 9–22, Federal Reserve Bank of Minneapolis.
- Prescott, E.C., 1986b, “Response to a Skeptic,” Quarterly Review 10, pp. 28–33 (Minneapolis: Federal Reserve Bank).
- Rotemberg, J.J., 2003, “Stochastic Technical Progress, smooth Trends, and Nearly Distinct Business Cycles,” American Economic Review, Vol. 93, No. 5, pp. 1543–59.

### Empirical studies on technology shocks, hours, and employment
- Carlsson, M., 2000, “Measures of Technology and the Short-Run Responses to Technology Shocks: Is the RBC Model Consistent with Swedish Manufacturing Data?,” (unpublished, Uppsala, Sweden: University of Uppsala).
- Chang, Y. and J.H. Hong, 2003, “On the Employment Effects of Technology: Evidence from U.S. Manufacturing for 1958–96” (unpublished, Philadelphia: University of Pennsylvania).
- Christiano, L., M. Eichenbaum, and R. Vigfusson, 2003, “What happens after a Technology Shock?” (unpublished, Evanston: Northwestern University).
- ———, 2004, “The Response of Hours to a Technology Shock: Evidence Based on Direct Measures of Technology,” EEA Meeting Papers and Proceedings, Journal of the European Economic Association (forthcoming).
- Collard, F. and H. Dellas, 2002, “Technology Shocks and Employment” (unpublished, Toulouse, France: Université de Toulouse).
- Fernald, J., 2004, “Trend Breaks, Long Run Restrictions, and the Contractionary effects of a Technology Shock” (unpublished, Chicago: Federal Reserve Bank).
- Fisher, J., 2003, “Technology Shocks Matter,” (unpublished; Chicago: Federal Reserve Bank).
- Francis, N., 2001, “Sectoral Technology Shocks Revisited,” (unpublished manuscript; Bethlehem, Pennsylvania: Lehigh University).
- Francis, N.R., M.T. Owyang, and A.T. Thedorou, 2003, “The Use of Long Run Restrictions for the Identification of Technology Shocks,” Federal Reserve Bank of St. Louis Review, November-December, pp. 53–66.
- ———, 2004, “What explains the Varying Monetary Response to Technology Shocks in G-7 Countries?,” (unpublished, Bethlehem: Lehigh University).
- Francis, N. and Valerie Ramey, 2003a, “Is the Technology-Driven Real Business Cycle Hypothesis Dead? Shocks and Aggregate Fluctuations Revisited” (unpublished, San Diego: University of California).
- ———, 2003b, “The Source of Historical Economic Fluctuations: An Analysis using Long-Run Restrictions” (unpublished, San Diego: University of California).
- Pesavento, E. and B. Rossi, 2003, “Do Technology Shocks Drive Hours Up or Down: A Little Evidence from an Agnostic Procedure” (unpublished, Durham: Duke University).
- Uhlig, H., 2004, “Do Technology Shocks Lead to a Fall in Total Hours Worked?” Journal of the European Economic Association, forthcoming.
- Wen, Y., 2001, “Technology, Employment and the Business Cycle: Do Technology Shocks Explain Aggregate Fluctuations? A Comment”. Cornel University Working Paper No. 01-19 (Ithaca, NY: Cornell University).

### Monetary policy, nominal rigidities, and identification methods
- Boivin, J. and M, Giannoni, 2003, “Has Monetary Policy Become More Effective?” NBER Working Paper 9459 (Cambridge: National Bureau of Economic Research).
- Bullard, J. and K. Mitra, 2002,”"Learning About Monetary Policy Rules”, Journal of Monetary Economics, Vol. 49, No. 6, pp. 1105–129.
- Calvo, G., 1983, “Staggered Prices in a Utility Maximizing Framework,” Journal of Monetary Economics, Vol. 12, No. 3, pp. 383–98.
- Christiano, L., M. Eichenbaum, and C. Evans, 2003, “Nominal Rigidities and the Dynamic Effects of a Shock to a Monetary Policy”, Journal of Political Economy, forthcoming.
- Erceg, C.J., L. Guerrieri, and C. Gust, 2004, “Can Long Run Restrictions Identify Technology Shocks?,” Federal Reserve Board Working Paper 792.
- Erceg, C.J. and A. Levin, 2003, “Imperfect Credibility and Inflation Persistence,” Journal of Monetary Economics, Vol. 50, No 4, pp. 915–44.
- Lubik, T. and F. Schorfheide, 2003, “Testing for Indeterminacy: An Application to U.S. Monetary Policy,” American Economic Review, Vol. 94, no. 1, pp. 190-217.
- Lubik, T. and F. Schorfheide, 2004, “Do Central Banks Respond to Exchange Rates? A Structural Investigation?” (unpublished, Philadelphia: University of Pennsylvania).
- Rabanal, P., 2003, “The Cost Channel of Monetary Policy: Further Evidence for the United States and the Euro Area,” IMF Working Paper 03/149 (Washington: International Monetary Fund).
- Romer, C., and D. Romer, 1989, “Does Monetary Policy Matter? A Test in the Spirit of Friedman and Schwartz,” NBER Macroeconomics Annual 1998, Bernanke, Ben and Julio Rotemberg, eds., Cambridge: MIT Press, pp. 63–129.
- Rotemberg, J.J., and M. Woodford, 1999, “Interest Rate Rules in an Estimated Sticky Price Model,” in Monetary Policy Rules, J.B. Taylor (ed.), Chicago: University of Chicago Press.
- Taylor, J.B., 1993, “Discretion versus Policy Rules in Practice,” Carnegie-Rochester Series on Public Policy, Vol. 39, pp. 195–214.
- Galí, J, J. D. López-Salido, and J.Vallés, 2003, “Technology Shocks and Monetary Policy: Assessing the Fed's Performance,” Journal of Monetary Economics, Vol. 50, No. 4, pp. 723–43.
- Galí, J., 1999, “Technology, Employment, and the Business Cycle: Do Technology Shocks Explain Aggregate Fluctuations?,” American Economic Review, Vol. 89, No. 1, pp. 249–71.
- Galí, J., and Mark Gertler, 1999, “Inflation Dynamics: A Structural Econometric Analysis,” Journal of Monetary Economics, Vol. 44, No. 2, pp. 195–222.
- Galí, J., 2003, “New Perspectives on Monetary Policy, Inflation, and the Business Cycle,” in Advances in Economics and Econometrics, Volume III, edited by M. Dewatripont, L. Hansen, and S. Turnovsky (Cambridge: Cambridge University Press).
- Galí, J., 2004, “On the Role of Technology Shocks as a Source of Business Cycles: Some New Evidence,” Journal of the European Economic Association, forthcoming.
- McGrattan, E., 1999, “Predicting the Effects of Federal Reserve Policy in a Sticky-Price Model,” FRB Working Paper 598 (Minneapolis: Federal Reserve Bank).
- McGrattan, E., 2004, “Comment to Galí and Rabanal”, in Mark Gertler and Ken Rogoff (eds.), NBER Macroeconomics Annual 2004, in press.

### Econometric and time-series methodology
- Baxter, M. and R. G. King, 1999, “Measuring Business Cycles: Approximate Band-Pass Filters for Economic Time Series,” The Review of Economics and Statistics, 81(4), pp. 575–93.
- Bils, M. and P. J. Klenow, 2002, “Some Evidence on the Importance of Sticky Prices,” NBER Working Paper 9069 (Cambridge: National Bureau of Economic Research).
- Cooley, T. F. and M. Dwyer, 1998, “Business Cycle Analysis Without Much Theory: A Look at Structural VARs,” Journal of Econometrics, Vol. 83, No. 1, pp. 57–88.
- Cooley, T. F. and E.C. Prescott, 1995, “Economic Growth and Business Cycles,” Frontiers of Business Cycle Research (Princeton: Princeton University Press).
- Hamilton, J., 1994, Time Series Analysis, Princeton University Press.
- Hoover, K., and S.J. Perez, 1994, “Post Hoc Ergo Procter Once More: An Evaluation of ‘Does Monetary Policy Matter?’ in the Spirit of James Tobin,” Journal of Monetary Economics, Vol. 34, Issue 1, pp. 47–74.
- Ingram, B., N. Kocherlakota, and N.E. Savin, 1994, “Explaining Business Cycles: A Multiple-shock Approach,” Journal of Monetary Economics, Vol. 34, Issue 3, pp. 415–28 (December).
- Kwiatkowski, D., P.C.B. Phillips, P. Schmidt and Y. Shin, 1992, “Testing the Null Hypothesis of Stationarity Against the Alternative of a Unit Root,” Journal of Econometrics, Vol. 54, pp. 159-178.
- Lettau, M. and H. Uhlig, 2000, “Can Habit Formation be Reconciled with Business Cycle Facts?,” Review of Economic Dynamics, Vol. 3, No. 1, pp. 79–99.
- Marcet, A., 2004, “Overdifferencing VAR's is OK” (unpublished, Barcelona, Spain: Universitat Pompeu Fabra).
- Fernández-Villaverde, J. and J.F. Rubio-Ramírez, 2004, “Comparing Dynamic Equilibrium Economies to Data: A Bayesian Approach,” Journal of Econometrics, forthcoming.
- Smets, F. and R. Wouters, 2003a, “An Estimated Stochastic Dynamic General Equilibrium Model of the Euro Area,” Journal of the European Economic Association, Vol 1 (5), pp. 1123–75.
- Smets, F. and R. Wouters, 2003b, “Shocks and Frictions in U.S. Business Cycles: A Bayesian DSGE Approach” (unpublished; Frankfurt: European Central Bank).
- Uhlig, H., 1999, “A Toolkit for Analyzing Nonlinear Dynamic Stochastic Models Easily,” in Computational Methods for the Study of Dynamic Economies, edited by Ramon Marimon and Andrew Scott (Oxford: Oxford University Press).

### Price stickiness, costs, and micro foundations
- Calvo, G., 1983, “Staggered Prices in a Utility Maximizing Framework,” Journal of Monetary Economics, Vol. 12, No. 3, pp. 383–98.
- Hall, R.E,. 1988, “The Relation Between Price and Marginal Cost in U.S. Industry,” Journal of Political Economy 96, pp. 921–47.
- McGrattan, E., 1994, “The Macroeconomic Effects of Distortionary Taxation,” Journal of Monetary Economics, Vol. 33, No. , pp. 573–601.
- Marchetti, D.J., and F. Nucci ,2004, “Price Stickiness and the Contractionary Effects of Technology Shocks,” European Economic Review, forthcoming.
- Calvo, G., 1983, “Staggered Prices in a Utility Maximizing Framework,” Journal of Monetary Economics, Vol. 12, No. 3, pp. 383–98.
- Hall, R.E., 1988, “The Relation Between Price and Marginal Cost in U.S. Industry,” Journal of Political Economy 96, pp. 921–47.

### Fiscal policy, measurement, and other foundations
- Cummins, J. and G. Violante, 2002, “Investment-Specific Technical Change in the United States (1947–2000),” Review of Economic Dynamics, Vol. 5, No. 2, pp. 243–84.
- Gordon, R., 1990, “The Measurement of Durable Goods Prices” (Chicago: University of Chicago Press).
- Jones, J.B., 2002, “Has Fiscal Policy Helped Stabilize the U.S. Economy?,” Journal of Monetary Economics, Vol 49, No. 4, pp. 709–46.
- Ramey, V., and M.D. Shapiro, 1998, “Costly Capital Reallocation and the Effects of Government Spending,” Carnegie-Rochester Conference Series on Public Policy, Vol. 48, pp. 145–194.
- Shea, J., 1998, "What Do Technology Shocks Do?," NBER Macroeconomics Annual 1998, Bernanke, Ben and Julio Rotemberg, eds., Cambridge: MIT Press, pp. 275-310.
- Stock, J., and M.W. Watson, 1999, “Business Cycle Fluctuations in U.S. Macroeconomic Time Series,” in J.B. Taylor and M. Woodford (eds.), Handbook of Macroeconomics, Volume 1A, pp. 3–64 (see also National Bureau of Economics Research Working Paper 6528).
- Solow, R., 1957, “Technical Change and the Aggregate Production Function,” Review of Economics and Statistics, Vol. 39, pp. 312–20.
- Orphanides, A. and J. Williams, 2002, “Robust Monetary Policy Rules with Unknown Natural Rates,” Brookings Papers on Economic Activity, Vol. 2002, pp. 63–118 (Washington: Brookings Institutions).
- Kiley, M.T., 1997, “Labor Productivity in U.S. Manufacturing: Does Sectoral Comovement Reflect Technology Shocks?” (unpublished, Washington: Federal Reserve Board).
- Francis, N., 2001, “Sectoral Technology Shocks Revisited,” (unpublished manuscript; Bethlehem, Pennsylvania: Lehigh University).
- Franco, F., and T. Philippon, 2004, “Industry and Aggregate Dynamics” (unpublished, Cambridge: Massachussets Institute of Technology).

*Source: _wp04234 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04234.pdf_