估算政治壓力對美聯儲的影響：一種采用新數據的敘事方法.pdf

《估算政治壓力對美聯儲的影響：一種采用新數據的敘事方法.pdf》由會員分享，可在線閱讀，更多相關《估算政治壓力對美聯儲的影響：一種采用新數據的敘事方法.pdf（51頁珍藏版）》請在三個皮匠報告上搜索。

1、estimating the effects of political pressure on the fed:a narrative approach with new dataThomas DrechselUniversity of Maryland,NBER,CEPRNovember 2024motivationRenewed attention to political dimension of U.S.monetary policyTrumps pressure to ease affected markets(Bianchi et al.,2023)Empirical resear

2、ch:Cross-country,e.g.Alesina and Summers(1993)Estimated regime-switching models,e.g.Bianchi and Ilut(2017)This paper:Identifies shocks to political pressure on the Fed over timeQuantifies effects on inflation and other macro variables1/18motivationRenewed attention to political dimension of U.S.mone

3、tary policyTrumps pressure to ease affected markets(Bianchi et al.,2023)Empirical research:Cross-country,e.g.Alesina and Summers(1993)Estimated regime-switching models,e.g.Bianchi and Ilut(2017)This paper:Identifies shocks to political pressure on the Fed over timeQuantifies effects on inflation and

4、 other macro variables1/18motivationRenewed attention to political dimension of U.S.monetary policyTrumps pressure to ease affected markets(Bianchi et al.,2023)Empirical research:Cross-country,e.g.Alesina and Summers(1993)Estimated regime-switching models,e.g.Bianchi and Ilut(2017)This paper:Identif

5、ies shocks to political pressure on the Fed over timeQuantifies effects on inflation and other macro variables1/18this paperNewly collected archival dataPersonal interactions between U.S.Presidents and Fed Officials 19332016Amount,length,type of interactions construct long time seriesNarrative ident

6、ificationExploit variation from Nixon and Johnson administrationsSVAR with narrative sign restrictions “political pressure shocks”2/18preview of findingsWhen the President pressures the Fed to ease monetary policy.1.Inflation and inflation expectations increase gradually and strongly2.Little to no e

7、ffect on other macro variables(inflation bias)3.Transmission is different from typical monetary policy expansion50%as much pressure as Nixon,for six months,raises U.S.price level by 8%3/18preview of findingsWhen the President pressures the Fed to ease monetary policy.1.Inflation and inflation expect

8、ations increase gradually and strongly2.Little to no effect on other macro variables(inflation bias)3.Transmission is different from typical monetary policy expansion50%as much pressure as Nixon,for six months,raises U.S.price level by 8%3/18data construction and descriptive evidencedata sourceDaily

9、 calendars of U.S.Presidents provided by Presidential LibrariesAvailable online or on site quality variesWith RAs,collect all interactions of Presidents with Fed Officials 1933-20164/18example of presidential scheduleTFE#MITE HGUSELOCATIQTHE WHITE HOUSEWASHINGTON,D.C.-THE DAILY DIARY OF PRESIDENT JI

10、MMY CARTERIDATE I MO.Clay.Yr.:JULY 19,1980TIMEDA15:30 a.m.SATURDAY 1IITIME3 72-4ux-TV 1-I-YIL2d:iII!5:30 R The President received a wake up call from the White Housesignal board operator.1 6:06 1The President went to the Oval Office.IIt1 8:oo t 8:01 P The President talked with the First Lady.I:8:08I

11、1 8:15The President returned to the second floor Residence.The President and the First Lady had breakfast.8:31I!g:10The President returned to the Oval Office.The President participated in a photo opportunity with:Mrs.Carol Anderson,Plains,GeorgiaJill Anderson,daughterTim Lawson,Americus,GeorgiaMrs.T

12、im(Peggie)LawsonWendi Lawson,daughterKeeli Lawson,daughterLorri Lawson,daughterThe President went to the South Grounds.The.President motored from the South Grounds to the Shoreham.Hotel.The President was greeted by:Clifford L.Alexander,Secretary of the ArmyRobert McIntosh,General Manager,Shoreham Ho

13、telThe President,escorted by Secretary Alexander,went to theVIP Room.The President met with:Secretary AlexanderMrs.Clifford L.(Adele)AlexanderGen.Edward C.Meyer,Chief of Staff,U.S.ArmyMrs.Edward C.(Carol)MeyerLt.Col.Matt Urban,U.S.Army and recipient ofthe Medal of HonorMrs.Matt(Jennie)UrbanJennifer

14、Urban,daughterKris Johnson,daughterMs.Sophia Rockwell,governessThe President participated in a ceremony to present the Medalof Honor to Lt.Col.Urban.Members of the pressTHE DAILY DIARY OF PaEsm&NT JIMMY CMTERILXATI 3NDATE r%io.Zay.Yr.,SHOREHAM HOTELWASHINGTON,D.C.SULY 19,1980TlM EDAY9:29 a.m.SATURDA

15、YI#ITSMEI1 9:309:32Ig:44III6I9:49II9:54po:15i1I10:3710:40r10:44105210:5g.11:02i11:02i!III93IIIIIIj 9:53/j4Ipo:35Igo:38IfIIiI1$0:57IIIill:50ill:50;11:50ill:50ill:50II!IHONE.KTIVITYThe President went to the offstage announcement area.The President went to the podium inside the Regency Ballroom.The Pre

16、sident addressed approximately 1,200 guests attendingthe presentation.The President presented the Medal of Honor to Lt.Col.Urban.The President returned to his motorcade.He was escorted bySecretary Alexander.The President motored from the Shoreham Hotel to the SouthGrounds of the White House.The Pres

17、ident returned to the Oval Office.The President met with Chairman of the Board of Governors ofthe Federal Reserve System,Paul A.Volcker.The President talked with the First Lady.The President telephoned his daughter,Amy Carter.The callwas not completed.The President went to the doctors office.The Pre

18、sident talked with Amy Carter.The President returned to the Oval Office.Enroute,hegreeted:Richard I.Queen,released U.S.HostageHarold Queen,father,resident of Lincolnville Beach,MaineMrs.Harold(Jeanne)QueenAlexander Queen,brotherWarren M.Christopher,Deputy Secretary of State.The Presidential party we

19、nt to the Oval Office.The President met with:Mr.Queen Mr.and Mrs.Harold QueenMr.Alexander QueenDeputy Secretary ChristopherThe First LadyII.1IIff5/18example of presidential schedule(zoomed in)5/18basic statistics857 personal interactions from 1933 to 2016Average length 53 minutes92%with Fed chair,8%

20、with other Fed Officials36%are 1-on-1 interactions11%are on Saturdays or Sundays16%happen in social settings,e.g.dinnerEnormous variation through time.6/18president-fed interaction count through time19401950196019701980199020002010020406080100Annual number of interactionsRooseveltTrumanEisenhowerKen

21、nedyLBJNixonFordCarterReaganBushClintonBush(Jr.)ObamaClinton:6 interactions in 8 years;Nixon:160 interactions in 5.5 yearsSecond half of 1971:Nixon speaks with Burns on average every 5 days7/18identification strategynarrative identificationChallenge:personal interactions do not necessarily capture p

22、olitical pressureSolution:find increases in President-Fed interactions that.took place for political reasons only,i.e.is plausibly exogenous.arguably impacted monetary policyExploit two historical episodes in SVARMain specification:Nixons pressure on Burns in 1971Additional variation:LBJs pressure o

23、n Martin in 19678/18nixons appointment of arthur burnsNixon during Arthur Burns swearing-in ceremony in 1970:“I respect his independence.However I hope that independently he willconclude that my views are the ones that should be followed.”9/18some quotes from arthur burns diary“I am convinced the Pr

24、esident will do anything to be reelected.”(March 1971)“I watched his face,as he spoke,with a feeling of dismay;for his features becametwisted and what I saw was uncontrolled cruelty.”(July 1971)“I got a stern letter from the President urging me start expanding the moneysupply and predicting disaster

25、 if this didnt happen.”(Oct 1971)“President at this meeting again expressed his concern about the money supply.Ireminded him that I was looking after that properly.”(Nov 1971)10/18narrative identification:corroborating evidenceBefore Nixon re-election(Feb 1970-Nov 1972)After Nixon re-election(Dec 19

26、72-July 1974)-6-4-20246810percentage pointsFed Funds Rate(cumulative change)Romer-Romer shock(cumulative)voting patterns11/18definition of political pressure shockLong quarterly data(Ramey and Zubairy,2018)+President-Fed interactionsNarrative sign restrictions(Antolin-Diaz and Rubio-Ramirez,2018)Var

27、iableTraditional sign restrictionsNarrative restrictionsPresident-Fed interaction count+1971:Q3,1971:Q4Log GDP deflator+3-month T-bill rateLog real GDPLog government expendituresNom.Deficit/Nom.GDPSVAR different from event study:can detect shock anywhere in samplepp vs.monetary shocks12/18resultsimp

28、ulse response functionsother samplesother shocksPresident-Fed Interactions 0 9182736-2024681012GDP Deflator 0 9182736-5051015%Tbill Rate 0 9182736-4-3-2-1012p.p.Real Government Expenditure 0 9182736quarters-30-20-10010203040%Real GDP 0 9182736quarters-15-10-5051015%Deficit per GDP 0 9182736quarters-

29、50510p.p.13/18historical variance decompositionshock seriesother decompsJan-36Jan-46Jan-56Jan-66Jan-76Jan-86Jan-96Jan-06Jan-16-4-2024681012Contribution of PP shockGDP Deflator,YoY log diff,%Evidence of political pressure shock affecting 1970s inflation,less elsewhere14/18effects on inflation expecta

30、tions0123456-4-20246p.p.Political Pressure Shock Mean inflation expectation0123456-4-20246p.p.Monetary Policy Shock Mean inflation expectation0123456-4-20246p.p.Generic 1971 Shock Mean inflation expectation0123456years-1-0.500.511.52p.p.Dispersion of inflation expectation0123456years-1-0.500.511.52p

31、.p.Dispersion of inflation expectation0123456years-1-0.500.511.52p.p.Dispersion of inflation expectationPolitical pressure shock raises inflation expectations and disagreement15/18effects on inflation expectationsReproduced with permission of the copyright owner.Further reproduction prohibited witho

32、ut permission.INSIDE THE MARKET:Wall Street Fears Whiplash From Nixon-Burns CollisionSchonberger,Ernest ALos Angeles Times(1923-1995);Feb 28,1971;ProQuest Historical Newspapers:Los Angeles Timespg.J1Pressure is definitely observed by the publicsystematic16/18irfs to pol.pressure shock identified fro

33、m nixon+lbjPresident-Fed Interactions 0 9182736-2024681012GDP Deflator 0 9182736-5051015%Tbill Rate 0 9182736-4-3-2-1012p.p.Real Government Expenditure 0 9182736quarters-30-20-10010203040%Real GDP 0 9182736quarters-15-10-5051015%Deficit per GDP 0 9182736quarters-50510p.p.Similar inflation response,t

34、ighter posterior bands17/18conclusionconclusionNovel data and narrative identification political pressure shocksQuantitative resultsPresident acts 50%as bad as Nixon for 6 months price level rises by 8%Benefits of central bank independence often highlighted using cross-country data;I provide support

35、ing evidence for the US through timeI make the new data and estimated shocks available on my website18/18references IAbrams,B.A.(2006):“How richard nixon pressured arthur burns:Evidence from the nixon tapes,”Journal ofEconomic Perspectives,20,177188.Alesina,A.and L.Summers(1993):“Central bank indepe

36、ndence and macroeconomic performance:somecomparative evidence,”Journal of Money,credit and Banking,25,151162.Antolin-Diaz,J.and J.F.Rubio-Ramirez(2018):“Narrative Sign Restrictions for SVARs,”AmericanEconomic Review,108,280229.Bianchi,F.,R.Gomez-Cram,T.Kind,and H.Kung(2023):“Threats to central bank

37、independence:High-frequency identification with Twitter,”Journal of Monetary Economics.Bianchi,F.and C.Ilut(2017):“Monetary/fiscal policy mix and agents beliefs,”Review of economicDynamics,26,113139.DeLong,J.B.(1997):“Americas peacetime inflation:the 1970s,”in Reducing inflation:Motivation andstrate

38、gy,University of Chicago Press,247280.Hetzel,R.L.(1998):“Arthur Burns and inflation,”FRB Richmond Economic Quarterly,84,2144.Meltzer,A.H.(2009):“A History of the Federal Reserve,Volume 2,Book 2,1970-1986,”Chicago/London.Ramey,V.A.and S.Zubairy(2018):“Government spending multipliers in good times and

39、 in bad:evidencefrom US historical data,”Journal of political economy,126,850901.Romer,C.D.and D.H.Romer(2004):“Choosing the Federal Reserve Chair:Lessons from History,”Journal of Economic Perspectives,18,129162.appendix slidessummary stats:1on1 vs otherInteraction is 1on1Other people are present010

40、0200300400500600Number of interactionsTotal duration(hours)backweekend interactions:count and length(in h)19401950196019701980199020002010051015RooseveltTrumanEisenhowerKennedyLBJNixonFordCarterReaganBushClintonBush(Jr.)Obama1940195019601970198019902000201005101520RooseveltTrumanEisenhowerKennedyLBJ

41、NixonFordCarterReaganBushClintonBush(Jr.)Obamabacksocial interactions:count and length(in h)194019501960197019801990200020100246810RooseveltTrumanEisenhowerKennedyLBJNixonFordCarterReaganBushClintonBush(Jr.)Obama194019501960197019801990200020100510152025RooseveltTrumanEisenhowerKennedyLBJNixonFordCa

42、rterReaganBushClintonBush(Jr.)Obamabackcomparison with martin(2015 eer)back1940195019601970198019902000201001020304050607080RooseveltTrumanEisenhowerKennedyLBJNixonFordCarterReaganBushClintonBush(Jr.)ObamaMeetings with Fed ChairEarlier data by Martin(2015)Cover longer time period and calendars that

43、are only physically available(Bush)Include wider set of Fed OfficialsCollect interaction characteristics:length,social nature,weekday,etc.Going to higher than annual frequencyMost importantly,use formally for identificationmore information about burns diarybackThe diary only became public in 2008Ent

44、ries cover meaningful interactions and/or when Burns was worried or angryFor example,large gap in entries before November 1970;it is a good guess thatpressures start around thereBurns himself writes in 1973 that“August 15,1971 was a definite anddecisive turning point in the Presidents state of mind.

45、”Also,there is a big break between February and September 1972Burns stopped writing the diary altogether after Nixon resignedmore quotes from burns diaryback“He was still the emperor,and I should therefore toe the mark.”(July 1971)“There was little room for any doubt(.)that he was governed mainly,if

46、 notentirely,by a political motive;(.)that the kind of changes we were discussing(.)were essential for the campaign of 1972”(Aug 1971)“President called and asked me to come over within an hour.Hastily rearrangedmy schedule and spent an hour and quarter with him.”(Sep 1971)“What a dangerous game RN i

47、s playing”(Nov 1971)“Clearly he wants power(.)but I dont think he has any positive idea at all as towhat to do with this power.”(Sep 1972)narrative identification:corroborating evidenceBefore Nixon re-election(Feb 1970-Nov 1972)After Nixon re-election(Dec 1972-July 1974)05101520#of dissenting FOMC v

48、otesDissenting votes in favor of tighter policyDissenting votes in favor of easier policyback17/18burns and nixon actively discussed fed personnel“Some talk followed about the inadequacies of Hayes and Brimmer.And thePresident indicated he would do what he could to rid me of Brimmer.”(Jan 1972)Alfre

49、d Hayes:NY Fed President.Andrew Brimmer:Member of Federal ReserveBoard.“We talked about personnel problem at Fed.President knew that Rogers wouldnot have Brimmer for U.N.We talked about an ambassadorship;Presidentthought that Sweden preferably Austria would be possibilities.I expressedstrong doubt a

50、bout Brimmer accepting an African post.”(Feb 1972)backdifferent views on whether burns gave in to pressureAccounts that agree with the view I portray here:Abrams(2006),Meltzer(2009)Different,at at least more nuanced,views:DeLong(1997),Hetzel(1998),Romer and Romer(2004)Main argument is that Burns had

51、 a“nonmonetary view of inflation”These accounts were provided before Burns diary was made public in 2008One has to acknowledge:1970s stagflation was a pretty new challengebackdifferent views on whether burns gave in to pressureFOMC voting behavior at least hard to reconcile with the“nonmonetary”view

52、being the standard view of the timeRomer-Romer easing shock:policy easing was nonsystematic!Burns diary makes clear there was a multidimensional discussion between Burnsand Nixon,touching on fiscal policy,restructuring the intl.monetary system,etc.Burns cared deeply about some of those issuesConceiv

53、able that increase in monetary supply was part of a“give and take”Even if Burns did not cave,expectations of economic agents could have beenchanged by Nixons behavior to the degree that it was publicly observableBurns diary makes clear that Press was well aware of the pressures(see next slide)See al

54、so Bianchi,Gomez-Cram,Kind,and Kung(2023)backpolitical pressure vs.monetary policy shocksAssume one period;:inflation,i:interest rates,x:President-Fed interactions=ii+xx+i=i+ixx+ix=x+xii+xRewrite second equation:i=i+ixx1ixxi+ixx+i1ixxi“Romer-Romer”:regress i on and get residual m=ixx+i1ixxiIf Presid

55、ent has no effect on Fed,then ix=0 and m=iGoal is to separately uncover xwith data on x and appropriate restrictionsTransmission is different from i,in particular if x=0back17/18public knowledge of nixons pressureSome quotes from Burns diary about the publics perception“The confrontation reported or

56、 predicted by the Press did not come off.”“(.)the White House children ought to stop the dangerous game of feeding gossipand fabrication to the Press.”“Recently,a journalist came to see me and told me that some White Houseoperatives(specifically,Haldeman and Shultz)had their bayonets out for me.”bac

57、kirfs over 1933-2016 sample,light restrictionsbackPresident-Fed Interactions 0 9182736-2024681012GDP Deflator 0 9182736051015%Tbill Rate 0 9182736-3-2-1012p.p.Real Government Expenditure 0 9182736quarters-20-100102030%Real GDP 0 9182736quarters-10-50510%Deficit per GDP 0 9182736quarters-4-20246p.p.S

58、tronger effect when interest rate unrestricted price level 7%higher after 10 yearsirfs over 1933-1987 samplebackPresident-Fed Interactions 0 9182736-2024681012GDP Deflator 0 9182736-5051015%Tbill Rate 0 9182736-3-2-1012p.p.Real Government Expenditure 0 9182736quarters-30-20-100102030%Real GDP 0 9182

59、736quarters-15-10-50510%Deficit per GDP 0 9182736quarters-6-4-20246p.p.In pre-Greenspan sample:negative effect on activity,government spending and deficitPolitical pressure not“successful”?popularityirfs over 1933-2008 samplebackPresident-Fed Interactions 0 9182736-2024681012GDP Deflator 0 9182736-2

60、024681012%Tbill Rate 0 9182736-3-2-1012p.p.Real Government Expenditure 0 9182736quarters-30-20-10010203040%Real GDP 0 9182736quarters-10-50510%Deficit per GDP 0 9182736quarters-4-20246p.p.Stopping before 2008 gives similar to stopping in 2016irfs over 1952-2016 samplebackPresident-Fed Interactions 0

61、 9182736-2024681012GDP Deflator 0 9182736-50510%Tbill Rate 0 9182736-4-3-2-1012p.p.Real Government Expenditure 0 9182736quarters-10-505%Real GDP 0 9182736quarters-505%Deficit per GDP 0 9182736quarters-2-10123p.p.Starting after 1952 gives different resultsVariation from before the Treasury-Fed accord

62、 appears to be important for the resultsnixons popularityback1969197019711972197319741975203040506070Approval,%Narrative restrictionAppointment of BurnsElectionSource:Gallup Pollsirfs to standard monetary policy shock 1933-1987backPresident-Fed Interactions 0 9182736-2-101234GDP Deflator 0 9182736-3

63、-2-10123%Tbill Rate 0 9182736-1.5-1-0.500.5p.p.Real Government Expenditure 0 9182736quarters-10-5051015%Real GDP 0 9182736quarters-2-101234%Deficit per GDP 0 9182736quarters-2-1012p.p.In pre-Greenspan sample,monetary easing increases real activityOpposite of political pressure shock in same sampleCo

64、nfirms that the transmission of the two shocks is very differentirfs to generic 1971 inflationary shockbackGDP Deflator 0 9182736-6-4-20246%Tbill Rate 0 9182736-3-2-101p.p.Real Government Expenditure 0 9182736-20-100102030%Real GDP 0 9182736quarters-50510%Deficit per GDP 0 9182736quarters-20246p.p.E

65、ffect on price level smaller and much less persistent(for roughly similar rate reduction)Highlights that personal interaction data key to my resultsestimated shocks over 1933-2016 sampleJan-35Jan-45Jan-55Jan-65Jan-75Jan-85Jan-95Jan-05Jan-15-3-2-10123456Huge spike in the 1970s,much less volatile afte

66、r the mid-1980sbackhistorical variance decomposition:interaction variableJan-35Jan-45Jan-55Jan-65Jan-75Jan-85Jan-95Jan-05Jan-15-5051015202530Contribution of PP shockPresident-Fed Interactionsbacksystematic evidenceback1940195019601970198019902000201005101520Share of articles about Fed pressure(%)RooseveltTrumanEisenhowerKennedyLBJNixonFordCarterReaganBushClintonBush(Jr.)ObamaSystematic search of NY Times,Wash Post,WSJ