The Possible Unemployment Cost of Average Inflation below a Credible - PDF document

The Possible Unemployment Cost of Average Inflation below a Credible Target Lars E.O. Svensson Sveriges Riksbank and Stockholm University May 16, 2012 1 Issue Average inflation below target 1996-2011  Average inflation expectations close to target 1996-2011  If inflation expectations stuck at target when average inflation  deviates from target, non-vertical long-run Phillips curve? Higher average unemployment from lower average inflation?  Bias in estimates of sustainable unemployment rate based on  historical averages? Conclusions for the future?  2

CPI inflation, CPI inflation expectations, and unemployment 1996-2011 12 CPI inflation 1yr inflation expectations 10 2yr inflation expectations Unemployment 8 6 4 2 0 -2 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 3 5-year moving averages: CPI inflation expectations close to 2 %, CPI inflation below 2 % 4.0 3.5 CPI inflation, 5yr moving average 1yr inflation expectations, 5yr moving average 3.0 2yr inflation expectations, 5yr moving average 2.5 2.0 1.5 1.0 0.5 0.0 00 01 02 03 04 05 06 07 08 09 10 11 4

Before 1996: High CPI inflation expectations (Aragon) 12 10 8 CPI inflation Inflation next 2 years 6 Inflation next 5 years Unemployment 4 2 0 -2 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 5 First few years: Inflation target not credible, tight monetary policy, and high unemployment e ( u u *) π − π = − γ − e π B * 2 π = π = u u* 6

Inflation target gradually becomes credible * 2 C B e π = π = π = u=u* 7 From 1996: Inflation expectations stuck at 2 %, but monetary policy still tight: Inflation too low, and unemployment too high * 2 C e π = π = D 1 .4 π = u* u 8

Estimate short-run Phillips curve, compute long-run Phillips curve 2 ( 2) ( 2) u π − = β + β π − + β π − + β + ε t 0 1 t 1 2 t 2 3 t t − − Coefficient Estimate Std. Error t-Statistic Prob. β 0 1.835275 0.528245 3.474290 0.0010 β 1 1.127351 0.121112 9.308356 0.0000 β 2 –0.500917 0.110455 –4.535029 0.0000 β 3 –0.283695 0.080984 –3.503117 0.0010 Newey-West lag 4, R 2 0.84, adj R 2 0.83, S.E. 0.51, DW 2.12 2 u / (1 ) / (1 ) π − = γ − γ γ = − β − β − β γ = β − β − β 0 0 0 1 2 3 1 2 π − 2 = 4.92 − 0.76 u 9 The long-run Phillips curve, 1998Q1-2011Q4 3 2 CPI inflation - 2, percentage points 1 0 -1 -2 -3 -4 5 6 7 8 9 10 Unemployment, percent 10

The unemployment cost of average inflation below a credible target ”Near-rational” inflation expectations?  Akerlof-Dickens-Perry (2000): For average inflation close  to zero, a significant fraction of agents disregard inflation; behave as if inflation expectations are zero Here, for average inflation close to 2 %, a significant  fraction of agents disregard deviation from 2 %; behave as if inflation expectations are 2 % Non-vertical Phillips curve applies for average inflation not  too fare from 2 % (± 1 %?) 11 The unemployment cost of average inflation below a credible target 1997-2011 average CPI inflation 1.4 %  Average inflation expectations about 2 %  Downward-sloping long-run Phillips curve  π t – 2 = 4.92 – 0.76 u t 0.6 p.p. lower inflation gives 0.6/0.76 = 0.8 p.p. higher  unemployment on average during 1997-2011 Robustness?  12

The unemployment cost of average inflation below a credible target Point estimate 0.79 p.p.  95 % conf. interval 0.79 ± 0.29 = [0.50 1.08] p.p.  99 % conf. interval 0.79 ± 0.39 = [0.40 1.18] p.p.  3.6 3.2 2.8 2.4 2.0 beta_0 1.6 1.2 0.8 0.4 0.0 -.6 -.4 -.2 .0 .2 .4 .6 0.79 + 0.6*(1 - beta_1 - beta_2)/beta_3 13 CPI inflation and unemployment, 1976-2011 Long-run Phillips curve, 1998-2011 16 12 1976 CPI inflation, percent 8 1993Q4 4 2011Q3 1998Q1 0 1 2 3 4 5 6 7 8 9 10 11 12 Unemployment,percent 14

Earlier start, 19 97 Q1-2011Q4 Flatter, higher unemployment cost 3 2 CPI inflation - 2, percentage points 1 0 -1 -2 -3 -4 5 6 7 8 9 10 11 12 Unemployment, percent 15 Later start, 19 99 Q1-2011Q4 Not much steeper, not much lower unemployment cost 3 2 CPI inflation - 2, percentage points 1 0 -1 -2 -3 -4 5 6 7 8 9 10 Unemployment, percent 16

Consider CPIXF inflation and gap to Riksbank long-term unemployment 12 CPI inflation CPIXF inflation 10 Unemployment Riksbank long-term unemployment 8 6 4 2 0 -2 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 17 With CPIXF inflation, 1998Q1-2011Q4 Flatter curve, similar unemployment cost 1.5 1 CPIXF inflation - 2, percentage points 0.5 0 -0.5 -1 -1.5 -2 5 6 7 8 9 10 Unemployment, percent 18

With Riksbank unemployment gap, higher cost, possibly overestimate of long-term unemployment 3 2 Inflation - 2, percentage points 1 0 -1 -2 -3 -4 -2 -1 0 1 2 3 Unemployment gap, percentage points 19 Revised Riksbank unemployment gap and long-run Phillips curve, 1998Q1-2011Q4 3 2 CPI inflation - 2, percentage points 1 0 -1 -2 -3 -4 -2 -1 0 1 2 3 Revised unemployment gap, percentage points 20

Revised Riksbank long-term unemployment gap, 12 CPI inflation Unemployment 10 Riksbank long-term unemployment Riksbank long-term unemployment, revised 8 6 4 2 0 -2 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 21 With lagged unemployment Flatter curve, higher unemployment cost 3 2 CPI inflation - 2, percentage points 1 0 -1 -2 -3 -4 5 6 7 8 9 10 Unemployment, percent 22

Other countries Average inflation in other countries?  CA average CPI 1996-2011 2.02 UK average RPIX 1995-2003 2.48 UK average CPI 2004-2007 2.01 UK average CPI 2008-2011 3.39 US average core CPI 2000-2011 2.02 US average core PCE 2000-2011 1.86 23 Canada, 1997Q1-2011Q4 Average inflation on target 2 % 3 2 CPI inflation - 2, percentage points 1 0 -1 -2 -3 5 6 7 8 9 10 Unemployment, percent 24

US, 2000Q1-2011Q4 (Fuhrer 2011) Average inflation 2 % 1.5 1 0.5 Core CPI inflation - 2, p.p. 0 -0.5 -1 -1.5 -2 -2.5 -2 -1 0 1 2 3 4 5 Unemployment gap, percent 25 Conclusions for the future? Swedish (Prospera) inflation expectations not rational  ”Near rational”?  Stuck at 2 % for average inflation not too far from 2 %? Stable inflation expectations of 2 per cent good:  Easier to stabilize unemployment without too much variation in inflation Important to hold average inflation close to 2 per cent  Too low average inflation can entail large real economic  costs Better with price-level targeting, average inflation targeting  over a longer period? 26

Government bill (1997/98:40, p. 1):  “ without prejudice to the objective of price stability , [the Riksbank] should support the objectives of general economic policy with the purpose to achieving sustainable growth and high employment .” 27 28