Total and Short-term Unemployment and the Recent Behavior of Prices and Wages

Background Background (cont.) • FRBNY staff analysis: considerable slack remains • Motivation to explore these issues: “missing in the economy deflation puzzle” • Labor market flows • “Standard” Phillips Curve models → very low inflation or deflation in the aftermath of the Great Recession • This assessment is an important factor in FRBNY • Compensation growth also should have been lower staff outlook • A restraining factor for price- and wage-inflation • One proposed explanation: Short-term/long-term unemployment distinction • However, recent research has focused on two • Long-term unemployed (unemployment duration > 27 issues: weeks): limited impact on price/wage inflation • Extent of slack in labor market • Short-term unemployment near historical average: less • Magnitude of restraining effect on inflation from slack than implied by conventional measures conventional measure(s) of labor market slack

Unemployment Rates by Duration Illustration of “Standard” Approach: Gordon (2013) Percent Percent 12 12 • Price-inflation (PCE deflator) Phillips curve model: Correlation Between Short- • Lagged inflation terms term and Long-term 10 Unemployment 10 • Unemployment gap based on: 1976—2007 0.63 • Total unemployment rate 1976—2013 0.28 8 8 • Short-term unemployment rate Total Unemployment • Supply shocks: 6 6 • Food and energy inflation • Relative import price inflation 4 Short-term 4 • Trend productivity acceleration/deceleration variable Unemployment Long-term • Other features: 2 Unemployment 2 • 1960:Q1-2013:Q1 sample period 0 0 • Joint estimation of time-varying NAIRU 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012 4 Source: Bureau of Labor Statistics Total and Short-term Unemployment and the Recent Behavior of Prices and Wages

Evaluation of Short-term vs. Long-term Distinction Percent Model-estimated NAIRUs Percent • Out-of-sample forecast performance (often during 7 7 NAIRU—Total post-2007 period) Unemployment 6 6 • Comparison of forecast errors using short-term and total unemployment gaps 5 5 NAIRU—Short-term • Within-sample estimates of the coefficient(s) on Unemployment the unemployment gaps over different sample 4 4 periods 3 3 • Are the estimates more stable using the short-term

unemployment gap? 2 2

1 1

0 0 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005 2009 2013 Source: Gordon 2013 Model-estimated Unemployment Gaps Percentage point Percentage point Annualized Quarterly PCE Inflation: Out of 6 6 Percent Sample Forecasts, 2007-Present Percent 5 5 8 8 Forecast using 4 4 6 Short-term 6 Unemployment 3 Total 3 Gap Unemployment 4 4 2 Gap 2 Short-term 2 2 1 Unemployment 1 Gap 0 0 0 0 Forecast Using -2 -2 -1 -1 Total Unemployment -4 Gap -4 -2 -2

-3 -3 -6 -6

-4 -4 -8 -8 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005 2009 2013 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: Gordon 2013 Source: Gordon 2013 Total and Short-term Unemployment and the Recent Behavior of Prices and Wages

Phillips Curve Stability and Forecast Contributions Summary of Other Studies in Short-term Unemployment Gap Model • Is the short-term unemployment rate a better Sum of unemployment coefficients with estimation through various dates indicator of labor market slack? • Empirical evidence at aggregate level is mixed:

1996Q4 2006Q4 2013Q1 Total Short-term Total Short-term Total Short-term • Very favorable -0.473 -0.741 -0.465 -0.726 -0.314 -0.641 • Ball and Mazumder (2011), Krueger, Cramer and Cho (2014), Macro Advisers (2014) • Slightly favorable • Feroli (2014), Hooper, Luzzetti and Slok (2014), Forecast Decompositions (Averages over period) Linder, Peach and Rich (2014), Watson (2014) 2008Q1-2009Q4 2010Q1-2013Q3 Imports -0.06 0.01 • Not convincing Food-energy -0.20 0.26 • Hatzius and Stehn/Mericle (2014) Unemployment -0.78 -0.76 Lagged inflation 2.38 1.76

Source: Gordon 2013 Summary of Other Studies (cont.) Discussion Questions • Analysis at aggregate level may be problematic • What is your current view concerning the short- • Little precedent for current labor market environment term/long-term unemployment debate? • High correlation between short- and long-term • What is your forecast for price- and wage inflation unemployment series prior to 2007 over the near- and medium-term horizons? • Additional checks for robustness needed • Some have claimed that labor market conditions • Alternative approach: use disaggregated data may be more closely linked to services inflation • Kiley (2014) – 24 large U.S. metropolitan areas than goods inflation. Do you believe this distinction • Short- and long-term unemployment exert equal is helpful in discussions of inflation and the downward pressure on price inflation inflation outlook? • Hooper, Luzzetti and Slok (2014) – State level data • Long-term unemployed more impactful in labor market as short-term unemployment gap closes