## 1. Okun’s Law: Output and Employment Gaps

## Source details

**Canonical URL:** [1. Okun’s Law: Output and Employment Gaps](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14216.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14216.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14216.pdf.json)

---

### I. Introduction — context and approach
- Unemployment in the OECD reached almost 7½ percent in March 2014, corresponding to about 46 million unemployed, 11 million more than in July 2008.
- Fiscal policy orientation: tilted toward supporting jobs by stimulating aggregate demand (IMF 2012).
- Channels from fiscal policy to labor market: output channel and direct channels (incentives, matching, training, employment protection adjustments).
- Okun’s Law benchmark: Okun (1962) found that a 1 percent increase in GNP corresponded to a 0.3 percentage point decrease in the unemployment rate; Okun’s coefficient varies across countries (examples: 0.8 for Spain; 0.15 for Japan).
- Paper’s empirical approach:
  - Calculate Okun’s coefficient for a panel of 34 OECD countries (1975–2012).
  - Estimate direct impacts of fiscal policy on employment gaps and conditional impacts via interaction between the output gap and fiscal policy.
  - Robustness: baseline analysis applied to unemployment gaps.

### II. Key empirical strategy and identification
- Baseline Okun specification:
  - Employment gap regressed on output gap with country fixed effects; trends obtained using Hodrick-Prescott filter with smoothing parameter 6.25.
- Extended specification (direct and conditional fiscal impacts):
  - Includes discretionary fiscal variables: total public expenditure, current primary expenditure, capital expenditure, wages and salaries, goods and services, social benefits, subsidies; and tax rates: CIT, PIT, VAT, SSC.
  - Interaction terms: fiscal variable * output gap capture how fiscal policy alters Okun’s relationship.
  - Interpretation:
    - β1 = Okun’s coefficient (employment gap response to output gap).
    - β2 = direct impact of fiscal instrument on employment gaps.
    - β3 = interaction effect (fiscal policy effect on responsiveness).
- Discretionary fiscal policy identification:
  - Residual from country-specific fiscal rule regression (fiscal balance on lagged balance, inflation, output gap, debt, time trend) used as proxy for fiscal activism.
- Controls and estimation:
  - Monetary policy stance controlled (central bank policy rate or alternative interest rates).
  - Labor market regulation index included; other institutional controls tested but not robust.
  - Within fixed effects estimator; bootstrap with 1,000 replications; correction for autocorrelation and heteroskedasticity.

### III. Baseline Okun’s Law estimates and core magnitudes
- Baseline Okun’s coefficient:
  - A 1 percent widening of the output gap leads to a 0.24 percentage point increase in the employment gap.
  - Reported Output Gap coefficient: 0.240*** (standard error (0.0232)); Observations: 538; R2: 0.464; Countries: 34.
- Employment vs. unemployment gaps:
  - Unemployment gap regression: Output Gap coefficient -0.335*** (standard error (0.0256)); Observations: 833; R2: 0.524; Countries: 34.

### IV. Fiscal expenditure effects on employment gaps (direct and conditional)
- Aggregate discretionary expenditure:
  - A 1 percent of GDP increase in discretionary expenditure expands employment gaps by 0.05 percentage points (Disc. Expenditure coefficient: 0.0515**; standard error (0.0230)).
  - Fiscal variables account for a limited fraction of variation in employment gaps, between 2–3 percent (R2 change reported when excluding fiscal variables on identical sample).
- Primary current expenditure:
  - Disc. Primary Current Expenditure coefficients: 0.0693*** and 0.0793*** (standard errors (0.0210), (0.0211)).
  - Interaction: Disc. Primary Current Expenditure*Output Gap: 2.668**, 2.915** (standard errors (1.270), (1.255)).
  - Numerical illustration: one percent of GDP more discretionary primary current spending raises Okun’s coefficient temporarily from 0.22 to 0.25; a 5 percent of GDP stimulus raises it to 0.35.
- Capital expenditure:
  - Disc. Capital Expenditure coefficients: 0.00426, 0.0528 (standard errors (0.0777), (0.0548)); interaction terms not significant — capital spending shows no significant short-run direct impact beyond aggregate demand.
- Monetary policy variable:
  - Monetary Policy Rate (-1) coefficients across specifications: 0.0135*, 0.0165*, 0.0184, 0.0201, 0.0180 (standard errors (0.00708), (0.00865), (0.0136), (0.0138), (0.0133)) — when significant, sign is positive.

### V. Employment multipliers by spending component (primary current discretionary expenditure)
- Estimated employment multipliers (primary current discretionary spending):
  - Goods and services: coefficient 0.28 (standard error (0.158)); when including selected items with sufficient observations coefficient can rise to 0.39 (text).
  - Wage bill (Disc. Wage & Salaries): coefficient 0.26 (standard error (0.130)).
  - Social benefits: coefficient 0.20 (standard error (0.0804)).
  - Subsidies: coefficient 0.14 (standard error (0.0713)).
- Interpretation and caveats:
  - Social benefits variable includes pensions (about one-third of social benefits) and likely captures assistance schemes, social services, and ad-hoc support programs rather than predominantly unemployment benefits.
  - Goods and services show the largest employment multiplier among components; a one percent of GDP spending on goods and services can lead to an increase in employment rates by up to 0.4 percentage points (textual statement).

### VI. Interaction effects and stability of Okun’s Law
- General finding:
  - Okun’s Law is largely stable; interaction terms for most spending components are insignificant.
- Notable exception — subsidies:
  - Disc. Subsidies*Output Gap significant and positive in some specifications (example coefficient 5.367*; standard error (2.883)).
  - Example numerical effect: spending one percent of GDP on discretionary subsidies increases Okun’s coefficient from 0.22 to 0.27 in some specifications.
  - At average positive output gaps, the spending multiplier of subsidies on employment reaches 0.21 (text).
- Asymmetry (positive vs negative output gaps):
  - Discretionary expenditure effects are smaller or offset during negative gaps:
    - Disc. Expenditure: 0.0967*** (standard error (0.0265)); Disc. Expenditure*Neg.Gap: -0.0864** (standard error (0.0418)).
    - Disc. Subsidies: 0.263** (standard error (0.0928)); Disc. Subsidies*Neg.Gap: -0.280*** (standard error (0.0897)) — implying subsidies’ positive effect on employment is almost entirely associated with positive output gaps; during recessions subsidies have no impact on employment beyond Okun’s Law.

### VII. Revenue-side (tax rates) effects on employment gaps and Okun’s coefficient
- Direct effects on employment gaps:
  - Social security contributions (SSC): a one percent increase in SSC lowers employment gaps by about 0.04 percentage points (SSC(-1): -0.0404**; standard error (0.0143)); effect can be 0.045 at average positive output gap (text).
  - Corporate income tax (CIT): a one percent increase lowers employment gaps by about 0.03 percentage points (CIT(-1): -0.0305**; standard error (0.0126)).
  - PIT and VAT: not significant in baseline employment estimates.
- Interaction effects (tax level and Okun’s coefficient):
  - Higher SSC rates reduce Okun’s coefficient: at average SSC rates Okun’s coefficient ≈ 0.24; at the 75th percentile of sample SSC rates Okun’s coefficient = 0.18 (text).
  - VAT: Okun’s coefficient is as low as 0.19 at the 75th percentile of sample VAT rates and reaches 0.34 at the 10th percentile of VAT sample rates (text).
- Sample and estimation notes:
  - Tax-rate estimations use smaller samples (Ilzetzki (2011) database: observations for 15 countries, 1981–2008); specific tables report Observations: 158; Countries: 13–14 depending on column.

### VIII. Unemployment gaps — robustness and comparison
- Discretionary spending effects on unemployment gaps (signs reversed relative to employment gaps and magnitudes differ):
  - Disc. Expenditure: -0.0580*** (standard error (0.0205)).
  - Disc. Primary Current Expenditure: -0.0707*** (standard error (0.0179)).
  - Disc. Goods & Services: -0.343*** (standard error (0.121)) — largest unemployment multiplier.
  - Disc. Social Benefits: -0.155* (standard error (0.0877)).
  - Disc. Subsidies: -0.117* (standard error (0.0613)).
  - Disc. Wage & Salaries: -0.179 (standard error (0.120)) — not significant.
- Tax policy and unemployment:
  - PIT(-1): 0.0107** in some specifications affecting unemployment gaps (standard error (0.00430)).
  - Other tax interactions with negative gap show mixed significance (examples: VAT(-1)*Negative Gap 0.0190* (standard error (0.00918)); SSC(-1)*Negative Gap 0.00960** (standard error (0.00430))).

### IX. Principal conclusions and policy implications
- Okun’s Law robustness:
  - Okun’s Law is robust across specifications but can be influenced by fiscal policy; a two percent of GDP increase in discretionary spending can shift Okun’s coefficient from around 0.2 towards 0.3 (textual example), mainly in expansions.
- Relative roles of fiscal instruments:
  - Current discretionary spending (especially goods and services and wages) has measurable positive short-run effects on employment.
  - Capital spending shows no significant short-run direct effect beyond aggregate demand.
  - Subsidies and social security contributions can alter the responsiveness of employment to output in expansions.
  - Revenue-side: cuts in corporate taxation (CIT) and social security contributions (SSC) can improve employment outcomes; PIT and VAT less robust in baseline employment regressions but VAT and SSC levels influence the size of Okun’s coefficient across countries.
- Magnitudes and practical considerations:
  - Spending one percent of GDP on goods and services can raise employment rates by up to 0.4 percentage points (textual statement).
  - Direct effects of specific tax rate changes on employment are small in magnitude, but average tax rate levels (VAT and SSC) strongly influence Okun’s coefficient.
- Policy recommendations and cautions:
  - Fiscal policy aimed directly at employment has smaller roles relative to policies that stimulate aggregate demand.
  - The optimal size of Okun’s coefficient is a political choice; tax design can influence employment sensitivity to output.
  - When using fiscal support for job creation, consider available fiscal space and potential distortionary effects given the relatively small estimated impacts.
  - Results concern short-term effects; some transmission channels of fiscal policy may take longer to materialize and warrant further research.

*Source: _wp14216 - 1. Okun’s Law: Output and Employment Gaps, PDF chapter.*

### 1. Okun’s Law: Output and Employment Gaps .......................................................................12

### 1. Okun’s Law: Output and Employment Gaps

### Section headings and page references
- 1. Okun’s Law: Output and Employment Gaps .......................................................................12
- 2. Employment Gaps and Discretionary Expenditure ..............................................................13
- 3. Employment Gaps and Primary Current Discretionary Expenditur ....................................15
- 4. Employment Gaps and Tax Rates ........................................................................................16
- 5. Employment Gaps, Positive and Negative Output Gaps, and Expenditure .........................17
- 6. Employment Gaps, Positive and Negative Output Gaps, and Tax Rates ............................18
- 7. Okun’s Law: Output and Unemployment Gaps ...................................................................19
- 8. Unemployment Gaps and Discretionary Expenditure .........................................................20
- 9. Unemployment Gaps and Tax Rates ....................................................................................21

### Figures listed
- Figure 1. Unemployment Rates in OECD Countries (2008–14) ...........................................................3
- Figure 2. Differences in Employment Rates Between March 2014 and September 2008 ....................4

*Source: _wp14216 - 1. Okun’s Law: Output and Employment Gaps, PDF chapter (canonical URL provided in source metadata).*

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

### _wp14216 - References .............................................................................................................

### I. INTRODUCTION
- Unemployment in the OECD reached almost 7½ percent in March 2014, corresponding to about 46 million unemployed, 11 million more than in July 2008.
- Fiscal policy has been tilted toward supporting jobs by stimulating aggregate demand (IMF 2012).
- Fiscal policy can affect labor demand and supply both via the output channel and more direct channels (e.g., incentives, matching, training, employment protection adjustments).
- Okun’s Law: Okun (1962) found that a 1 percent increase in GNP corresponded to a 0.3 percentage point decrease in the unemployment rate.
- Okun’s coefficient varies across countries (examples cited: 0.8 for Spain; 0.15 for Japan).
- The paper’s approach:
  - Calculate Okun’s coefficient for a panel of 34 OECD countries (1975–2012).
  - Measure direct impacts of fiscal policy on employment gaps and the conditional impact via interaction between the output gap and fiscal policy.
  - Apply baseline analysis to unemployment gaps as robustness check.
- Key summary findings (overview):
  - Discretionary current expenditure robustly affects employment gaps; capital spending has no significant short-run impact beyond aggregate demand.
  - Discretionary spending on goods and services has the largest effect among spending components, followed by the wage bill, social benefits, and subsidies.
  - Cutting corporate income tax (CIT) and social security contributions (SSC) rates can improve employment outcomes.
  - Okun’s coefficient is broadly stable against most policy changes, but subsidies and social security contributions can affect Okun’s relationship at times of positive output gaps.
  - Some spending components have direct, albeit small, impacts — implying fiscal interventions to reduce employment gaps play a minor role relative to policies stimulating aggregate demand.

### II. LITERATURE REVIEW
- Expenditure side:
  - Public spending generally has a positive effect on labor market outcomes, mainly through aggregate demand.
  - Wage bill affects employment directly via public sector employment.
  - Mixed empirical findings: some studies find government purchases and wage bill increases reduce private sector employment and investment (Lane and Perotti (2003); Alesina and others (2002)); others find positive effects on employment (Fatas and Mihov, 2001; Burnside and others, 2004; Cavallo, 2005; Gali and others, 2007; Monacelli and others, 2010).
  - Social benefits can weaken the link between labor supply and incomes and tend to reduce labor demand; cited estimates:
    - Duval and Bassanini: a 10 percent increase in unemployment benefits would increase unemployment by 1.2 percentage points.
    - Krueger and Meyer (2002): a 10 percent increase in unemployment benefits raises average duration of unemployment by around 5 percent.
  - Pension benefits increases tend to reduce labor force and employment.
- Revenue side:
  - Labor taxes (personal income tax and social security contributions) negatively affect employment by reducing labor supply and/or increasing labor costs for firms (Bassanini and Duval, 2006).
  - Elasticities cited:
    - Cahuc and Zylberberg (2004): price elasticity of labor demand close to about 1.
    - IMF (2012): elasticity of labor supply to real wages between 0.2 and 0.5 percent.
  - Corporate taxes can affect employment via reduced investment and production and through potential pass-through to lower wages.
  - Consumption taxes (VAT, excises) increase costs for consumption goods and can reduce real wages and labor supply; compositional shifts from labor to consumption taxes can boost labor demand (fiscal devaluation example).
  - Long-term employment effects of tax shifts depend on how much the burden moves away from labor income; adjustment can take time (De Mooij and Keen, 2012).

### III. EMPIRICAL STRATEGY
- Baseline Okun’s Law specification (employment gap on output gap):
  - Employment gap (deviation of current employment from trend) and output gap (deviation of actual output from potential).
  - Trend and potential obtained using Hodrick-Prescott filtering with smoothing parameter 6.25 (other parameters considered; results not sensitive).
  - Equation (1) specification summarizes employment gap = α0 + α1 * output gap + country fixed effects + error term.
- Extended specification to estimate direct and conditional fiscal impacts (equation (2)):
  - Includes vector of discretionary fiscal policy variables (࣮_it): total public expenditure, current primary expenditure, capital expenditure, spending on wages and salaries, goods and services, social benefits, other spending (notably subsidies), and tax rates (CIT, personal income tax, VAT, SSC).
  - Interaction term between fiscal variables and the output gap captures influence of fiscal policy on Okun’s relationship.
  - Interpretation:
    - β1 = Okun’s coefficient (how much employment gap changes with output gap).
    - β2 = direct impact of fiscal instrument on employment gaps, controlling for output gap.
    - β3 = interaction effect: if significant, fiscal policy alters how employment responds to output.
  - Marginal effects:
    - Change in employment gap with respect to output gap: β1 + β3 * fiscal variable (equation (3)).
    - Change in employment gap with respect to fiscal policy: μ0 + μ1 * output gap (equation (4)).
- Controls and estimation:
  - Controls for monetary policy stance (central bank policy rate; when unavailable money market rate or long-term bond yield).
  - Index of strictness in labor market regulations included but largely not significant; other institutional controls tested but estimates not robust and sample loss large.
  - Within fixed effects estimator used; bootstrap with 1,000 replications to estimate equation (2).
  - Autocorrelation and heteroskedasticity corrected; intercept included.
- Discretionary fiscal policy identification:
  - Discretionary fiscal policy computed as residual from country-specific fiscal rule regression of fiscal balance on lagged balance, inflation, output gap, debt, and time trend (equation (5)), following Fatas and Mihov (2003, 2006), Afonso and others (2010), Agnello and others (2013).
  - Residual taken as proxy for fiscal activism; tax-rate estimations control for expenditure changes to isolate tax effects.
- Data:
  - Panel of 34 OECD countries, 1975–2012.
  - Employment: OECD database.
  - Real GDP and public spending items: IMF World Economic Outlook (WEO).
  - Tax rates: Ilzetzki (2011) database (observations for 15 countries, 1981–2008).
  - Interest rates: IMF-International Financial Statistics (IFS).
  - Labor market regulation and components: World Economic Freedom dataset (WEF); employment protection legislation: OECD.

### IV. RESULTS
- Baseline Okun’s Law:
  - Estimated Okun’s coefficient: a 1 percent widening of the output gap leads to a 0.24 percentage point increase in the employment gap.
  - Table 1 summary statistics:
    - Output Gap coefficient: 0.240***
    - Standard error: (0.0232)
    - Observations: 538
    - R2: 0.464
    - Countries: 34
    - Estimation includes country fixed effects and intercept.
- Adding fiscal policy — expenditure side:
  - Total discretionary expenditure: a 1 percent of GDP increase in expenditure expands employment gaps by 0.05 percentage points.
  - Fiscal policies account for a limited fraction of variation in employment gaps, between 2–3 percent (R2 change reported when excluding fiscal variables on identical sample).
  - Effect driven mainly by current primary expenditure:
    - Current primary expenditure coefficient around 0.07 (0.08 when including capital spending simultaneously).
  - Capital spending is not significant in the short-run beyond aggregate demand effects.
  - Monetary policy variable, when significant, has a positive sign (interpreted as substitution effect from labor to capital prevailing over negative output effect).
  - Interaction term for primary current spending is significant and positive:
    - Implies that a change in output has a larger effect on employment when discretionary current spending is expansionary.
    - Numerical illustrations:
      - For one percent of GDP more discretionary primary current spending, Okun’s coefficient would increase temporarily from 0.22 to 0.25.
      - For a 5 percent of GDP stimulus, Okun’s coefficient would increase to 0.35.
  - Overall, changes in Okun’s coefficient due to interaction terms are relatively small.
- Summary implications from results:
  - Discretionary current spending (especially goods and services, wage bill, social benefits, subsidies) has measurable, but modest, direct effects on employment gaps.
  - Capital spending shows no significant short-run direct impact on employment gaps beyond aggregate demand.
  - Revenue-side measures: cutting CIT and SSC rates can improve employment outcomes (highlighted in Introduction summary).
  - Okun’s Law is largely stable across most fiscal policy changes; notable exceptions include subsidies and social security contributions at times of positive output gaps.
  - Overall inference: fiscal policy interventions aimed directly at employment have smaller roles relative to policies that stimulate aggregate demand.

*Source: _wp14216 - References .............................................................................................................*

### 0.10 and 0.12 (based on results shown in Table 2).

### _wp14216 - 0.10 and 0.12 (based on results shown in Table 2)

### Employment Gaps and Discretionary Expenditure
- Output Gap coefficients (various specifications): 0.242***, 0.236***, 0.222***, 0.220***, 0.219*** (standard errors: (0.0244), (0.0263), (0.0262), (0.0301), (0.0279)).
- Monetary Policy Rate (-1) coefficients: 0.0135*, 0.0165*, 0.0184, 0.0201, 0.0180 (standard errors: (0.00708), (0.00865), (0.0136), (0.0138), (0.0133)).
- Disc. Expenditure coefficient: 0.0515** (standard error (0.0230)); Disc. Expenditure*Output Gap: 1.378 (standard error (1.689)).
- Disc. Primary Current Expenditure coefficients: 0.0693***, 0.0793*** (standard errors (0.0210), (0.0211)); Disc. Primary Current Expenditure*Output Gap: 2.668**, 2.915** (standard errors (1.270), (1.255)).
- Disc. Capital Expenditure coefficients: 0.00426, 0.0528 (standard errors (0.0777), (0.0548)); Disc. Capital Expenditure*Output Gap: 1.747, 1.503 (standard errors (4.701), (4.377)).
- Observations across specifications: 519, 496, 400, 400, 400. R2: 0.472, 0.480, 0.430, 0.406, 0.435. Non-Fiscal R2 (where reported): ..., 0.470, 0.404, 0.404, 0.404. Countries: 34, 34, 29, 29, 29.
- Estimation includes country fixed effects and intercept. Standard errors in parentheses. Significance: *** p<0.01, ** p<0.05, * p<0.1.

### Employment multipliers by spending component (primary current discretionary expenditure)
- Goods and services: coefficient 0.28 (largest employment multiplier among components in Table 3 context; standard error reported later as (0.158) in Table 3 regressions).
- Wage bill (Disc. Wage & Salaries): coefficient 0.26 (standard error (0.130)).
- Social benefits: coefficient 0.20 (standard error (0.0804)).
- Subsidies: coefficient 0.14 (standard error (0.0713)).
- For goods and services, when including selected spending items with sufficient observations (including capital spending), the coefficient can rise to 0.39 (result reported in text).
- Interpretation notes: social benefits variable includes pensions (about one-third of social benefits) and reflects non-discretionary components; social benefits here likely capture assistance schemes, social services, and ad-hoc support programs rather than predominantly unemployment benefits.

### Interaction with Output Gap (stability of Okun’s Law)
- For all components except subsidies, the interaction term with the output gap is insignificant, indicating stability of Okun’s Law.
- For subsidies, Disc. Subsidies*Output Gap is significant and positive in some specifications: example coefficient 5.367* (standard error (2.883) in Table 3 panel) and in Table 2 interactions subsidy-related results indicate Okun’s coefficient increases from 0.22 to 0.27 when government spends one percent of GDP on discretionary subsidies.
- At average positive output gaps, the spending multiplier of subsidies on employment reaches 0.21 (textual statement).

### Selected regression outputs for components (Table 3)
- Output Gap: 0.218***, 0.219***, 0.218***, 0.221*** (standard errors (0.0357), (0.0356), (0.0297), (0.0352)).
- Disc. Wage & Salaries: 0.261** (standard error (0.130)); Disc. Wage & Salaries*Output Gap 8.004 (standard error (6.391)).
- Disc. Goods & Services: 0.280* (standard error (0.158)); Disc. Goods & Services*Output Gap 11.54 (standard error (9.396)).
- Disc. Social Benefits: 0.203** (standard error (0.0804)); Disc. Social Benefits*Output Gap 3.369 (standard error (3.417)).
- Disc. Subsidies: 0.141** (standard error (0.0713)); Disc. Subsidies*Output Gap 5.367* (standard error (2.883)).
- Observations reported: 309, 295, 387, 280. R2: 0.396, 0.391, 0.408, 0.401. Non-Fiscal R2: 0.377, 0.378, 0.391, 0.380. Countries: 23, 22, 26, 21.

### Tax rates and employment (Table 4 and related text)
- Significant negative effects on employment gaps:
  - Social security contributions (SSC): a one percent increase in SSC lowers employment gaps by about 0.04 percentage points; effect can be 0.045 at average positive output gap (textual).
  - Corporate income tax (CIT): a one percent increase lowers employment gaps by about 0.03 percentage points (textual).
- PIT and VAT rates: not significant on employment in baseline estimates (Table 4 summary).
- Interaction effects on Okun’s coefficient:
  - Higher SSC rates reduce Okun’s coefficient: at average SSC rates Okun’s coefficient ≈ 0.24; at the 75th percentile of sample SSC rates Okun’s coefficient = 0.18 (textual).
  - VAT: Okun’s coefficient is as low as 0.19 at the 75th percentile of sample VAT rates and reaches 0.34 at the 10th percentile of VAT sample rates (textual).
- Table 4 sample output highlights (examples):
  - Output Gap: 0.213**, 0.434***, 0.151*, 0.374*** (standard errors (0.0741), (0.0466), (0.0786), (0.0758)).
  - Monetary Policy Rate (-1): 0.0197*, 0.0207*, 0.0430*, 0.0174* (standard errors (0.0107), (0.0105), (0.0224), (0.00947)).
  - PIT (-1): -0.00785 (standard error (0.00505)).
  - VAT (-1)*Output Gap: -1.150*** (standard error (0.296)).
  - CIT(-1): -0.0305** (standard error (0.0126)).
  - SSC(-1): -0.0404** (standard error (0.0143)); SSC(-1)*Output Gap: -0.395* (standard error (0.199)).
  - Observations: 158 across columns; R2: 0.583, 0.609, 0.608, 0.640. Countries: 14, 14, 14, 13.

### Positive vs Negative Output Gaps (asymmetric effects)
- Dummy for negative output gaps introduced to test asymmetric fiscal effects.
- Main findings:
  - Disc. Expenditure: 0.0967*** (standard error (0.0265)); Disc. Expenditure*Neg.Gap: -0.0864** (standard error (0.0418)) — suggests discretionary expenditure effects are smaller or offset during negative gaps.
  - Disc. Wage & Salaries: 0.377*** (standard error (0.122)); Disc. Wage & Salaries*Neg.Gap -0.219 (standard error (0.200)).
  - Disc. Goods & Services: 0.447** (standard error (0.211)); Disc. Goods & Services*Neg.Gap -0.305 (standard error (0.250)).
  - Disc. Social Benefits: 0.251** (standard error (0.102)); Disc. Social Benefits*Neg.Gap -0.0930 (standard error (0.144)).
  - Disc. Subsidies: 0.263** (standard error (0.0928)); Disc. Subsidies*Neg.Gap -0.280*** (standard error (0.0897)).
- Interpretation: the positive effect of subsidies on employment is almost entirely associated with positive output gaps; during recessions, subsidies have no impact on employment beyond Okun’s Law.
- Observations across related specifications: 496, 400, 400, 309, 295, 387, 280. R2 range: 0.484 to 0.399. Countries: 34, 29, 29, 23, 22, 26, 21.

### Unemployment gaps: comparison and results
- Okun’s Law for unemployment gaps:
  - Unemployment Gap regression: Output Gap coefficient -0.335*** (standard error (0.0256)). Observations: 833. R2: 0.524. Countries: 34.
- Discretionary spending effects on unemployment gaps:
  - Disc. Expenditure: -0.0580*** (standard error (0.0205)).
  - Disc. Primary Current Expenditure: -0.0707*** (standard error (0.0179)).
  - Disc. Goods & Services: -0.343*** (standard error (0.121)).
  - Disc. Social Benefits: -0.155* (standard error (0.0877)).
  - Disc. Subsidies: -0.117* (standard error (0.0613)).
  - Disc. Wage & Salaries: -0.179 (standard error (0.120)) — not significant.
- Tax policy and unemployment:
  - PIT (-1): 0.0107** in some specifications affecting unemployment gaps (standard error (0.00430)).
  - Other tax interactions with negative gap show mixed significance (examples: VAT(-1)*Negative Gap 0.0190* (standard error (0.00918)); SSC(-1)*Negative Gap 0.00960** (standard error (0.00430))).
- Interpretation: spending multipliers are lower for unemployment than employment; goods and services show the strongest impact on unemployment with a multiplier of 0.34. Tax effects on unemployment appear to capture labor supply dynamics (PIT important).

### Conclusion and policy implications
- Okun’s Law is robust across specifications but can be influenced by fiscal policy:
  - A two percent of GDP increase in discretionary spending can shift Okun’s coefficient from around 0.2 towards 0.3 (textual example), and this influence occurs primarily in expansions.
- Fiscal instruments affecting labor market outcomes in the short run:
  - Current discretionary spending (especially goods and services and wages) has measurable positive effects on employment.
  - Cuts in corporate taxation (CIT) and social security contributions (SSC) positively affect employment gaps (reductions in rates associated with higher employment gaps).
  - Subsidies and increases in SSC can magnify the impact of fiscal policy on employment during positive output gaps.
- Magnitudes and practical considerations:
  - Spending one percent of GDP on goods and services can lead to an increase in employment rates by up to 0.4 percentage points (textual statement).
  - Direct effects of specific tax rate changes on employment are small in magnitude, but average tax rate levels (VAT and SSC) strongly influence the size of Okun’s coefficient across countries.
- Policy recommendations and cautions:
  - The optimal size of Okun’s coefficient is a political choice; fiscal policy—particularly tax design—can influence employment sensitivity to output.
  - When using fiscal support for job creation, consider fiscal space and potential distortionary effects given the relatively small estimated impacts.
  - These results concern short-term effects; some transmission channels of fiscal policy may take longer to materialize and warrant further research.

*Source: _wp14216 - 0.10 and 0.12 (based on results shown in Table 2).*

### References

### References

### Fiscal policy, employment, and labor markets
- Agnello, L., D. Furceri, and R. M. Sousa, 2013, “How Best to Measure Discretionary Fiscal Policy? Assessing its Impact on Private Spending,” Economic Modelling, Vol. 34, pp. 15–24.
- Afonso, A., L., Agnello, and D. Furceri 2010, “Fiscal Policy Responsiveness, Persistence and Discretion,” Public Choice, Vol. 145, No. 3, pp. 503–30.
- Alesina A., S. Ardagna, R. Perotti, and F. Schiantarelli, 2002, “Fiscal Policy, Profits, and Investment,” American Economic Review, Vol. 92, pp. 571–89.
- Ardagna, S., 2007, “Fiscal Policy in Unionized Labor Markets,” Journal of Economic Dynamics and Control, Vol. 31, pp. 1498–1534.
- Brückner, M., and E. Pappa, 2012, “Fiscal Expansions, Unemployment, and Labor Participation,” International Economic Review, Vol. 53, pp. 1205–28.
- De Mooij, and M. Keen, 2013, ‘Fiscal Devaluation’ and Fiscal Consolidation: The VAT in Troubled Times,” in Fiscal Policy After the Crisis, ed. by A. Alesina and F. Giavazzi, (Chicago: University of Chicago Press).
- Devries, P., J. Guajardo, D. Leigh, and A. Pescatori, 2011, “A New Action-Based Dataset of Fiscal Consolidations,” IMF Working Paper No. 11/128 (Washington: International Monetary Fund).
- Estevão, M. 2007, “Labor Policies to Raise Employment,” IMF Staff Papers, Vol. 54, No. 1.
- Kato, R. R., and H. Miyamoto, 2013, “Fiscal Stimulus in an Endogenous Job Separation Model,” International University of Japan Research Institute, Economics and Management Series EMS-2013-02 (Tokyo: UIJ Research Institute). Available at: http://www.iuj.ac.jp/research/workingpapers/EMS_2013_02.pdf.
- Monacelli T., Perotti R., and A. Trigari, 2010, “Unemployment fiscal multipliers,” Journal of Monetary Economics, 97, No. 5, pp. 531–53.
- Pappa, E., 2009, “The effects of fiscal shocks on employment and the real wage,” International Economic Review, Vol. 50, pp. 217–44.
- Tagkalakis, A., 2013, "The unemployment effects of fiscal policy: recent evidence from Greece," IZA Journal of European Labor Studies, Springer, Vol. 2, No. 1, pp. 1–32, December (Bonn: Institute for the Study of Labor).
- Vitek, F., 2010, “Output and Unemployment Dynamics during the Great Recession: A Panel Unobserved Components Analysis,” IMF Working Paper No. 10/185 (Washington: International Monetary Fund).

### Measurement, methodology, and datasets
- Ball, L., D. Leigh, and P. Loungani, 2013, "Okun's Law: Fit at Fifty?" NBER Working Papers No. 18668 (Cambridge: National Bureau of Economic Research).
- Brambor, T, W. R. Clark, and M. Golder, 2006, “Understanding Interaction Models: Improving Empirical Analysis,” Political Analysis, Vol.14, pp. 63–82.
- Burnside, C., M. Eichenbaum and J. Fisher 2004, “Fiscal Shocks and their Consequences,” Journal of Economic Theory, Vol. 115, No. 1, pp. 89–117.
- Card, D, J. Kluve, and A. Weber, 2010, “Active Labour Market Policy Evaluations: A Meta-Analysis,” The Economic Journal, Vol. 120 (November), pp. 452–77.
- Chen, J., P. Kannan, P. Loungani, and B. Trehan, 2011, “New Evidence on Cyclical and Structural Sources of Unemployment,” IMF Working Paper No. 11/106 (Washington: International Monetary Fund).
- Daveri, F., and G. Tabellini, 2000, “Unemployment, Growth and Taxation in Industrial Countries,” Economic Policy, Vol. 30, pp. 49–104.
- Finn, M., 1998, “Cyclical Effects of Government’s Employment and Goods Purchases,” International Economic Review, Vol. 39, No. 3, pp. 635–57.
- Gali J., J. D. Lopez-Salido, and J. Valles, 2003, “Understanding the Effects of Government Spending on Consumption,” European Central Bank Working Paper No. 339 (Brussels: European Central Bank).
- Ilzetki, E., and C.A. Vegh, 2008, “Procyclical fiscal Policy in Developing Countries: Truth or Fiction?” NBER Working Paper No. 14191 (Cambridge: National Bureau of Economic Research).
- Ramey, V. A., 2012, “Government Spending and Private Activity,” NBER Working Paper No. 17787 (Cambridge: National Bureau of Economic Research).
- Romer, C. D., and D. H. Romer, 2010, “The Macroeconomic Effects of Tax Changes: Estimates Based on a New Measure of Fiscal Shocks,” American Economic Review, Vol. 100, No. 3, pp. 763–801.

### Labor market institutions, social insurance, and structural unemployment
- Bassanini, A., and R. Duval, 2006, “Employment Patterns In OECD Countries: Reassessing the Role of Policies and Institutions,” OECD Social, Employment and Migration Working Papers No. 35 (Paris: Organisation for Economic Co-Operation and Development).
- Cahuc, P., and A. Zylberberg, 2004, Labor Economics (Cambridge: MIT Press).
- Disney, R., 2004, “Are Contributions to Public Pension Programmes a Tax on Employment?” Economic Policy, Vol. 19, No. 39, pp. 267–311.
- Gordon, R., 2010, “Okun’s Law and Productivity Innovations,” American Economic Review: Papers and Proceedings, Vol. 100, No. 2, pp. 11–5.
- Krueger, A.B., and B.D. Meyer, 2002, “Labor Supply Effects of Social Insurance,” NBER Working Paper No. 9014 (Cambridge: National Bureau of Economic Research).
- Nickell, S., 1998, “Unemployment: Questions and Some Answers,” The Economic Journal, Vol. 108, No. 448.
- Nunziata, L., 2002,”Unemployment, Labour Market Institutions and Shocks,” Nuffield College Working Papers in Economics 2002-W16 (Oxford: Nuffield College).
- Orlandi, F., 2012, “Structural unemployment and its determinants in the EU countries,” European Economy, Economic Papers, Vol. 455 (Brussels: European Commission).
- Phelps, E. S., 1994, Structural Slumps (Cambridge: Harvard University Press).
- Scarpetta, S., 1996, “Assessing the Role of Labour Market Policies and Institutional Settings on Unemployment: A Cross-Country Study,” OECD Economic Studies, Vol. 26, No. 1, pp. 43–98.

### Composition of fiscal policy, taxation, and current account effects
- Arnold, J., 2008, “Do Tax Structures Affect Aggregate Economic Growth? Empirical Evidence from a Panel of OECD Countries,” OECD Economics Department Working Paper No. 643 (Paris: Organisation for Economic Co-Operation and Development).
- Auerbach A. J., and L. J. Kotlikoff, 1987, Dynamic Fiscal Policy (Cambridge: Cambridge University Press).
- Cavallo, M., 2005, “Government Consumption Expenditures and the Current Account,” Federal Reserve Bank of San Francisco Working Paper 2005-03 (San Francisco: Federal Reserve Bank).
- De Mooij, and M. Keen, 2013, ‘Fiscal Devaluation’ and Fiscal Consolidation: The VAT in Troubled Times,” in Fiscal Policy After the Crisis, ed. by A. Alesina and F. Giavazzi, (Chicago: University of Chicago Press).
- Lane P.R., and R. Perotti, 2003, “The importance of composition of fiscal policy: evidence from different exchange rate regimes,” Journal of Public Economics, Vol. 87, pp. 2253–79.
- Romer, C. D., and D. H. Romer, 2010, “The Macroeconomic Effects of Tax Changes: Estimates Based on a New Measure of Fiscal Shocks,” American Economic Review, Vol. 100, No. 3, pp. 763–801.

### IMF, OECD, and major reports
- International Monetary Fund, 2010, World Economic Outlook: Rebalancing Growth, April (Washington).
- International Monetary Fund, 2012, “Fiscal Policy and Employment in Advanced and Emerging Economies,” Board Paper prepared by the Fiscal Affairs Department (Washington: International Monetary Fund).
- Organisation for Economic Co-operation and Development (OECD), 2009, “The Jobs Crisis: What Are the Implications for Employment and Social Policy,” OECD Employment Outlook (Paris).

*Source: _wp14216 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14216.pdf_