Wage formation, unemployment and business cycle in Latvia

Transcription

1 Wage formation, unemployment and business cycle in Latvia Ginters Buss (Latvijas Banka) Full paper at Ginters IDEAS 1 of 14

2 1. Taylor or Calvo-type wage frictions unsuitable for Latvia more than once a year once a year between 1 to 2 years every two years before less frequently than every two years never/not applicable Figure: Frequency of base wage change in Latvia (%) Source: Fadejeva and Krasnopjorovs (215). 2 of 14

3 2. Thus try & compare alternative wage formations alternating-offer bargaining (AOB) without exogenous wage rigidity (CET, 216 Econometrica; Hall & Milgrom 28 AER) Nash bargaining with Taylor-type frictions (Buss, 215; CTW) Nash bargaining without exogenous wage rigidity reduced-form bargaining rule simple, reduced-form wage rule in a full-sized, estimated open economy New-Keynesian model with financial accelerator as in Buss (215,216) & CTW. Nash bargaining threat: discontinue firm-worker relationship AOB bargaining threat: prolong bargaining process weaker relation b/w wages & outside option 3 of 14

4 3. AOB vs Nash bargaining sharing rules Sharing rule: J t = { β1 (V t U t ) β 2 γ b,t + β 3 (ϑ t b u t ), for AOB 1 η (V η t U t ), for Nash J t - firm surplus from having a worker V t - worker surplus from having a job U t - worker value of unemployment ϑ t - real price of intermediate good bt u - unemployment benefit β 1, β 2, β 3 - restricted functions of δ b, M M - # bargaining subperiods (even) δ b - prob of break-up γ b - cost of counter-offer for a firm η - worker bargaining power (Nash) 4 of 14

5 4. Reduced-form specifications Reduced-form sharing rule J t = ɛ 1 (V t U t ) ɛ 2 + ɛ 3 (ϑ t b u t ), ɛ-s unrestricted. Nests AOB and Nash sharing rules as special cases. AOB: ɛ 1 = β 1, ɛ 2 = β 2 γ b, ɛ 3 = β 3. Nash: ɛ 1 = (1 η)/η, ɛ 2 = ɛ 3 =. Simple, reduced-form wage rule ln( w t / w) = ι 1 ln( w t 1 / w) + ι 2 ln(l t 1 /L) + ι 3 ln (µ z,t /µ z ) + ι 4 ln ɛ t. w t - real wage scaled by unit-root technology trend L t 1 - employment in t-1 µ z,t - unit-root neutral technology growth ɛ t - stationary neutral technology growth ι 1, ι 2, ι 3, ι 4 - free parameters to be estimated. 5 of 14

6 5. Impulse responses AOB vs Nash Nom. interest rate (APP) Net worth (% dev.) Country risk premium shock Consumption (% dev.) Total hours (% dev.) -2-4 CPI inflation (APP) Investment (% dev.) Net 1exports/Y -3 (Lev.dev.) Real wage (% dev.) GDP (% dev.) Real exch. rate (% dev.) Spread (APP) Hours/employee (% dev.) Unempl. 1 rate 2 (Lev.dev.) Shadow wage, MPL(% dev.) Net f. assets/gdp(lev.dev.) AOB 95% AOB mean Nash Taylor Nash flex 6 of 14

7 6. 1-quarter ahead forecasts AOB vs Nash-Taylor AOB: no excess volatility in wage forecasts despite the absence of exogenous wage rigidity Output Total hours observed filtered, Nash Taylor filtered, AOB Real wage Unemployment rate, level of 14

8 7. 1-quarter ahead forecasts AOB vs Nash-Taylor Nash-Taylor w/o labor supply shock struggles to forecast vacancies note: AOB has no labor supply shock either Hiring rate (at data mean) Total # vacancies, qoq Separation rate (at data mean) Total # vacancies, level of 14

9 8. 1-quarter ahead forecasts, alternatives alternative specifications; AOB the best (reduced-form sharing rule in b/w AOB & Nash-flex) Output 5 Total hours observed filtered, Nash flex filtered, wage rule filtered, AOB Real wage Unemployment rate, level of 14

10 q forecasting performance w.r.t. RW better if < 1; AOB among the best Model AOB Nash Taylor Wage rule SVAR AOB vs Nash flex wage Distance 1Q 4Q 8Q 12Q measure CPI GDP CPI GDP CPI GDP CPI GDP RMSE DM p-val MAE DM p-val RMSE MAE RMSE MAE RMSE MAE real wage forecasts RMSE MAE of 14

11 Simulation: min wage in Latvia Minimum average wage in LV from 33%( ) to > 4% looks like a permanent level shift ( ) vs (29-216) minimum to average wage (from firm surveys), % minimum to average wage (from national accounts), % minimum wage, EUR (rhs) Source: Central statistical bureau of Latvia, likumi.lv and author s calculations 11 of 14

12 11. Simulation: min wage in Latvia (2) (Affected) firm share that considers a particular cost-correction action (due to min wage ) was important (21-215): action BG EE HU LT LV RO SI total fire hiring product prices other costs wages for others productivity Source: ESCB Wage Dynamic Network Simulation strategy: There is no min wage nor wage distribution in the model but permanently worker bargaining power s.t. firms wages 12 of 14

13 12. Simulation: min wage in Latvia (3) 3 endog re-actions: price, hiring, factor substitution potentially high long-term losses in employment & output Wage increase due to higher worker bargaining power Nominal wage(%dev) Real wage(%dev) Marginal cost(%dev) # vacancies(%dev) Employment(Lev.dev) CPI inflation(app) Real exports(%dev) Real consump(%dev) GDP(%dev) Export price infl(app) Real investm(%dev) Net worth(%dev) AOB Nash-Taylor of 14

14 13. Conclusion AOB is a decent alternative to Nash bargaining: can live w/o exogenous wage rigidity otherwise similar behavior Simulation of permanent min wage (relative to average): approximate by permanently worker barg.power that induces firms wages identifies 3 endogenous re-actions ( price, hiring, factor substitution) supported by surveys potentially high long-term losses in employment & output (about -3% for Latvia ( ) vs (29-216)) 14 of 14