M ake A m erica G reat: Lo ng-R un Im p a cts o f S h o rt-R u n P - - PowerPoint PPT Presentation

M ake A m erica G reat: Lo ng-R un Im p a cts o f S h o rt-R u n P u b lic Investm e nt

A le x a n d r K o p y to v and H a o tia n X ia n g The Wharton School M a y 9, 2018

M A G : L o n g - R u n Im p a c t s o f S h o r t- R u n P u b lic I n v e s t m e n t 1 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

£ + F o llo w D o n a ld J . T ru m p O @realDonaldTrump

T h e ro a d s a n d s id e w a lk s , a irp o rts a n d b rid g e s , a re p e rfe c t in D u b a i. E v e ry th in g lo o k s c le a n & s tro n g . In U .S ., e v e ry th in g is fa llin g a p a rt!

RETWEETS LIKES 71 ־ g£Q-.

• ,.* n [־=־, r־» V

I *

257 258 i i B £. X A

¥ \\ 3 1

7:33 AM -1 9 May 2014 134 257 V 258

E

H illa ry C lin to n O 3

±* F o llo w

v @HillaryClinton

In v e s tin g in o u r in fra s tru c tu re is a b o u t m o re th a n c re a tin g g o o d jo b s : it's a b o u t m a in ta in in g o u r s ta tu s a s th e w o r ld ’s e c o n o m ic s u p e rp o w e r.

RETWEETS LIKES r 4 ־ k. t v ~ ■OB ■ ■ i

598 1,335

7:06 PM -3 0 Nov 2015

♦ » 184 598 V 1.3K

< □ ►

4 i f ► 4 0 ^ 0־ =51■ ► 5 » ► ן

M A G : L o n g - R u n Im p a c ts o f S h o r t - R u n P u b lic I n v e s tm e n t 2 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

► L a rry S u m m e rs: “T h e issue n o w is n o t w h e th e r th e US sh o u ld in ve st [in in fra s tru c tu re ] m ore b u t w h a t th e p o lic y fra m e w o rk sh o u ld be .” ► Ben B e rn a n ke : “ I w o u ld th in k ... th a t in fra s tru c tu re s p e n d in g th a t im p ro ve s o u r roads, o u r bridges, o u r schools, and ... w o u ld p ro b a b ly be th e h ig h e s t-re tu rn fisca l a c tio n s in te rm s o f g e ttin g h ig h e r g ro w th .” ► P aul K ru g m a n : “ In fra s tru c tu re please.” ► Joseph S tig litz : “ In fra s tru c tu re alo n e c o u ld a b sorb trillio n s o f d o lla rs in in v e s tm e n t, n o t o n ly tru e in th e d e v e lo p in g w o rld , b u t also in th e US, w h ic h has u n d e rin ve ste d in its core in fra s tru c tu re fo r decades.” = ►

• s | = 0י a o־< □

M A G : L o n g - R u n Im p a c t s o f S h o r t- R u n P u b lic I n v e s t m e n t 3 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

M o tiv a tio n

► O n e o f th e fe w consensuses: A large-scale s h o rt-ru n g o v e rn m e n t in v e s tm e n t (G I) p ro g ra m ► C ru cia l t o u n d e rsta n d e c o n o m ic consequences p ro d u ce d by a large tra n s ito ry G I sh o ck ► C u rre n t m a c ro fra m e w o rk s : C o n sid e r sm a ll G I sh o cks in (a lm o s t) lin e a r R B C m o d e ls (B a x te r and K in g , 1993; Leepe r e t al., 2 0 1 0 ) ► Is th e re n o n -lin e a rity ? ־0 a oי =|

s

• ►

=

4 □ M A G : L o n g - R u n Im p a c t s o f S h o r t- R u n P u b lic I n v e s t m e n t 4 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

M o tiv a tio n

F acts: In te rs ta te H ig h w a y C o n s tru c tio n a ro u n d 1960s N o n -d e fe n se Gl to o u tp u t O u tp u t p e r c a p ita , lo g d e v ia tio n

1950 1955 1960 1965 1970 1975 1980 1985 1990

In v e s tm e n t p e r c a p ita , lo g d e v ia tio n

1950 1960 1970 1980 1990 2000 2010

C o n s u m p tio n p e r c a p ita , lo g d e v ia tio n

1950 1955 1960 1965 1970 1975 1980 1985 1990 1950 1955 1960 1965 1970 1975 1980 1985 1990

0, 0 ^ ( = 5 . ־׳ =

M A G : L o n g - R u n I m p a c ts o f S h o r t - R u n P u b li< n d H a o t ia n X ia n g

O vervie w

► W e c o n s tru c t a R B C m o d e l w ith tw o s ta b le ste a d y s ta te s (s.s.) ► M u ltip lic ity arises because public capital com plem ents private choice ► Small shocks have tem porary outcomes, while large ones can lead to steady state transitions ► Rare switches across steady states can generate medium term cycles as in the data ► A p p ly o u r m o d e l t o tw o case stu d ie s: ► Can the large GI in the 1960s explain S-shaped dynam ics o f the US economy? ► Can a large GI expansion trigger a recovery from the Great Recession?

= ►

• eI

= 0 , >ג0י4 □

M A G : L o n g - R u n Im p a c t s o f S h o r t- R u n P u b lic I n v e s t m e n t 6 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

L ite ra tu re Review

► R B C m o d e ls w ith tra n s ito ry s h o c k to GI ► Baxter and King (1993), Leeper, Walker and Yang (2010) ► E n d o g e n o u s g ro w th m o d e ls w ith p e rm a n e n t ch a n g e in GI ► Barro (1990), Futagami, Morita and Shibata (1993), Eicher and Turnovsky (2000) ► E m p iric a l an a lysis o f GI ► Public capital elasticity: Aschauer (1989), Bom and Ligthart (2013) ► Nonlinear impact of GI: Fernald (1999), Candelon, Colletaz and Hurlin (2013) ► Multiplier papers: Perotti (2004), Auerbach and Gorodnichenko (2012), Ilzetzki, Mendoza and Vegh (2013) ► R B C m o d e ls w ith m u ltip le ste a d y s ta te s ► Schaal and Taschereau-Dumouchel (2015), Cai (2016) ► M e d iu m te rm flu c tu a tio n s ► Comin and Gertler (2006), Anzoategui, Comin, Gertler, Martinez (2017)

< □ ►

4 & ► 4 = ► 4 ^ ►

• I

I =

• 0 0 , 0־

M A G : L o n g - R u n Im p a c t s o f S h o r t- R u n P u b lic I n v e s t m e n t 7 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

• I. M o d e l

I ■ s ►

• sl = ■

0.0

M A G : L o n g - R u n Im p a c t s o f S h o r t- R u n P u b lic I n v e s tm e n t A le x a n d r K o p y t o v a n d H a o tia n X ia n g

H ouseholds

► H o u se h o ld w ith G H H prefe rence s u b je c t t o d is to rtio n a ry in c o m e ta x and lu m p sum ta x (tra n s fe rs ): 1+V 1 ץ— Y 1 + v )

C t - k

1 - Y E ^ !3* m a x t s.t. C t = (1 - T ) ( W tLt + R tK t + n t ) - It - Tt K t+ 1 = (1 - 5k)K t + It >ג,0י =| s

• ►

=

< □

M A G : L o n g - R u n Im p a c t s o f S h o r t- R u n P u b lic I n v e s t m e n t 9 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

F irm s

► A c o n tin u u m o f s h o rt-liv e d firm s p ro d u ce u n d e r C -D te c h n o lo g y s u b je c t to n o n -c o n v e x te c h n o lo g y ch o ice u £ { H , L}:

n = m a x {n H , k l }

w h e re n H = m a x A ld/ k°k( K G) a w - W l - R k- f

k,/ s

• ---------------------v----------------------׳

high (H) utilization

n L = m a x A ld/k°k( K G) a - W l - Rk k,/ s

^ ✓ low (L) utilization ► A is a g g re g a te sh o ck: lo g A t + 1 = pA lo g A t + c t a c ^ , eA ~ N ( 0 ,1 ) ► D e fin e A n = ! h - n\_. F irm chooses H w h e n A n > 0

4 □ ► < g ► 4 = ► < 1 ►

• l | = •0 0,0־

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 1 0 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

N ( 0 , 1)

■ s ►

• si = OHO

► L a w o f m o tio n o f p u b lic c a p ita l: K t+ 1 = (1 - £g)K tG + Ct w h e re G l/o u tp u t ra tio g t = C t/Y t fo llo w s A R (1 ): g t+ 1 = (1 - pg)־ l + pggt + ag eg+ 1 , £g ► G ov co n su m e s a given fra c tio n o f o u tp u t:

Fiscal Rule

► F in a n cin g :

C l + C f = T Yt + Tt 4

□

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 11 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

M echanism s

► F ro m p u b lic c a p ita l to p riv a te c a p ita l K G t ^ K , L, Y t ^ K G t ► (1) C om plem entarity between private factors, ( K , L), and public capital, K G, in Cobb-Douglas production function ► (2) GI proportional to o u tp u t ► F ro m c a p ita l a c c u m u la tio n K , K G t to te c h n o lo g y a d o p tio n ► (3) A non-convex adoption cost = ►

• 3 1

= 0י a o

4

□

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 1 2 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

► D enote fra c tio n o f H firm s as m e [0 ,1 ] ► D efine A l־ l(m ss) as A 7 r(m ss) evaluated a t steady state w ith m = mss P ro p o s itio n

The model has two stable deterministic steady states mss e { 0 , 1 } if i)

A l־ l(m ss) is increasing in mss and ii) A f l ( l ) > 0 > A fl(O ).

C h a ra cte riza tio n

M u ltip le S te a d y S ta te s

► i) requires m oderate social IRS: 1 —

9k — > e t> ן0ך^

► ii) non-convex cost creates no deviatio n in centive and su pp o rts tw o stable steady states

Mass of firms operating high te ch n o lo g y,

C h a ra cte riza tio n

M u ltip le S te a d y S ta te s

M A G : L o n g - R u n I m p a c t s o f S h o r t - R u n P u b lic In v e s t m e n t 1 4 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

Public capital, K

• II. C a lib ra tio n and A ssessm ents

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic In v e s t m e n t A le x a n d r K o p y t o v a n d H a o tia n X ia n g

P a ra m e triza tio n I

Parameter Value Source

Preferences

Risk aversion Labor elasticity Time discounting Y = 1.0 1/v = 3.33

@

= 0.951/4 Log utility Higher end of macro estimates 0.95 annually

Production function

Labor share Capital share Public capital elasticity

e! = 0.56 ek = 0.24 a = 0.15

Basu and Fernald (1997) Basu and Fernald (1997) Bom and Ligthart (2013)

Depreciation rates

Private capital Public capital

& k

= 1 - 0.91/4

S g

= 1 - 0.921/4 10% annually 8% annually, Leeper et al (2010)

Fiscal Spending Rule

Government consumption Transfers Government investment

g C

= 0.235 z = 0.060

g1

= 0.041 Postwar US data Postwar US data Postwar US data < □ ► 4

S ►

<-= ► < ■ e ►

• = I

= 'OQ.O

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 1 6 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

P a ra m e te riza tio n II

Parameter Value Source

GI shock

Persistence Standard deviation of shocks

P g

= 0.967

(7 g

= 0.0011 Postwar US data Postwar US data

TF P

Persistence Standard deviation of shocks PA = 0.94

a

a = 0.008 Output persistence Output volatility

Technology adoption

Fixed cost Scaling up parameter

f = 0.0051

w = 1.02 Frequency of transitions Distance between steady states >ג,0י = | s

• ►

=

4

□

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 1 7 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

IRFs: GI S hocks a t s.s. L

GI shock, % o f o u tp u t T F P shock, lo g d e v ia tio n M easured T F P , lo g d e v ia tio n O u tp u t, log d e v ia tio n In v e s tm e n t, lo g d e v ia tio n C o n s u m p tio n , lo g d e v ia tio n H ours, lo g d e v ia tio n P u b lic c a p ita l, lo g d e v ia tio n M ass o f H firm s

4 □ ► 4 [fp ► 4 Y ► 4 f=

►

• E I

= 0 ,0 0־

A le x a n d r K o p y t o v a n d H a o tia n X ia n g M A G : L o n g - R u n Im p a c t s o f S h o r t- R u n P u b lic I n v e s t m e n t 1 8 / 2 9

IRFs: T F P S hocks a t s.s. L

GI shock, % o f o u tp u t T F P shock, lo g d e v ia tio n M easured T F P , lo g d e v ia tio n O u tp u t, log d e v ia tio n In v e s tm e n t, lo g d e v ia tio n C o n s u m p tio n , lo g d e v ia tio n M ass o f H firm s H ours, lo g d e v ia tio n P u b lic c a p ita l, lo g d e v ia tio n

20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 200

1 ► 3 1 ■ 5

• 0 0 ,0

M A G : L o n g - R u n I m p a c t s o f S h o r t- R u n P u b lic In v e s tm e n t A le x a n d r K o p y t o v a n d H a o tia n X ia n g

M e d iu m -T e rm Cycles

Medium-term cycle, 0-200 qtr High frequency component, 0-32 qtr Medium frequency component, 32-200 qtr Data Model Data Model Data Model Output 4.23 4.46 (2.99,6.46) 2.30 2.05 (1.59,2.61) 3.61 3.91 (2.28,6.05) Consumption 3.34 3.62 (2.36,5.37) 1.30 1.43 (1.11,1.80) 3.12 3.28 (1.92,5.20) Hours 3.75 3.42 (2.31,4.87) 1.78 1.57 (1.23,1.97) 3.41 3.00 (1.76,4.59) Investment 12.35 10.61 (7.24,15.64) 8.05 5.86 (4.33,7.92) 9.35 8.65 (5.09,14.24) TFP 2.46 2.17 (1.57,2.44) 1.52 1.16 (0.93,1.44) 1.95 1.80 (1.10,2.65) ► In fre q u e n t s w itc h e s betw een ste a d y s ta te s g e n e ra te m e d iu m fre q u e n c y flu c tu a tio n s (C o m in and G e rtle r, 2 0 0 6 )

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 2 0 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

IV . Case S tu d ie s I ■ s ►

• sl = ■

0.0

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic In v e s t m e n t A le x a n d r K o p y t o v a n d H a o tia n X ia n g

C an GI sh o cks in th e 1960s e xp la in S -d y n a m ic s o f th e m a c ro e co n o m y? I ■ s ►

• sl = 0 0.0

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic In v e s t m e n t A le x a n d r K o p y t o v a n d H a o tia n X ia n g

1960s H ighw ay C o n stru ctio n

T ra n s itio n P a th s

► E x tra c t GI shocks fro m th e d a ta : 1 9 60Q 1 - 1 9 7 2 Q 4 ► P ro d u c tiv ity sh o cks backed o u t th ro u g h th e residual a p p ro a ch (m e a su re d T F P ) Output, log deviation Consumption, log deviation Investment, log deviation

M A G : L o n g - R u n Im p a c ts o f S h o r t - R u n P u b lic I n v e s t m e n t n d H a o t ia n X ia n g

In v e s tm e n t, lo g d e v ia tio n

1960 1965 1970 1975 1980 1985 1990

M ass o f H firm s

1960s H ighw ay C o n stru ctio n

D e c o m p o s itio n : Roles o f T F P and GI S ho cks

0.3 F 0 .2 ■ 0.1 ■

• 0.1 ■
• 0 .2 [_

_ _

1955

1 —

0.5 ■ 1955

O u tp u t, lo g d e v ia tio n

0.15 0.1 0.05

• 0.05

1955 1960 1965 1970 1975 1980 1985 1990 0.2 0.15 0.1 0.05

• 0.05

1955 1960 1965 1970 1975 1980 1985 1990

P u b lic c a p ita l, lo g d e v ia tio n

— / / '

. . . — — -

"A nother republican president, D w ight D. Eisenhower, initiated the last tru ly great national infrastructure program - the building o f the interstate highway system. The tim e has come for a new program o f national rebuilding.”

• D o n a ld J. T ru m p ’s S peech t o th e Congress, Feb 28, 2017

= ►

• s | = 0י a o־< □

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 2 5 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

W o u ld a large GI s h o ck have help ed th e US e c o n o m y to re co ve r fro m th e crisis? = ►

• s | = 0י a o־< □

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 2 6 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

T h e S low Recovery

T ra n s itio n P a th s

► E x tra c t GI shocks fro m th e d a ta : 2 0 0 7 Q 4 - 2 0 1 7 Q 2 . ► P ro d u c tiv ity sh o cks backed o u t th ro u g h th e residual a p p ro a ch (m e a su re d T F P )

C o n s u m p tio n , lo g d e v ia tio n O u tp u t, lo g d e v ia tio n

• -- U d I d
• --M o d e l -

־% — ^ ^ -»

In v e s tm e n t, lo g d e v ia tio n

0.4 0.2

• 0 .2
• 0.4

« a > < .. = ? >

• g|= >

0,0

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

• 1

□ ► 4 g

M A G : L o n g - R u n Im p a c ts o f S h o r t - R u n P u b lic I n v e s t m e n t n d H a o t ia n X ia n g

T h e S low Recovery

D e c o m p o s itio n : Roles o f T F P and GI S ho cks

► Y e llo w lines: a c o u n te rfa c tu a l increase in GI s ta rtin g a t 2 0 0 9 Q 3 o f ro u g h ly 1 trillio n 2 0 0 9 d o lla rs Investment, log deviation Output, log deviation Mass of H firms

• ז"/־

y־1V Public capital, log deviation

2017 2016 2015 2014 2013 2012 2011 2010 2009 2008

00,0

• =

5 1

• ׳4 5 4 ►5 4 '5 ► □ 4•

M A G : L o n g - R u n Im p a c ts o f S h o r t - R u n P u b lic I n v e s t m e n t n d H a o t ia n X ia n g

C onclusion

► W e d o c u m e n t S -shaped d y n a m ic s o f US m a c ro e c o n o m y a ro u n d th e c o n s tru c tio n o f th e In te rs ta te H ig h w a y S yste m in th e 1960s ► W e b u ilt an R B C m o d e l w ith m u ltip le s.s. to c a p tu re such a n o n -lin e a r e ffe c t o f a large GI ► In a “depressed” econom y, a te m p o ra ry GI p ro g ra m can c re a te a p e rm a n e n t sca le -u p ► In fre q u e n t s w itc h e s betw een ste a d y s ta te s can g e n e ra te m e d iu m te rm cycle s as in th e p o stw a r US d a ta ► E xogeno us p ro d u c tiv ity sh o cks play c ru c ia l roles = ►

• s | = 0י a o־< □

M A G : L o n g - R u n Im p a c ts o f S h o r t- R u n P u b lic I n v e s tm e n t 2 9 / 2 9 A le x a n d r K o p y t o v a n d H a o tia n X ia n g

Gross M a cro Series

O u tp u t, lo g d e v ia tio n C o n s u m p tio n , lo g d e v ia tio n

1950 1955 1960 1965 1970 1975 1980 1985 1990 1950 1955 1960 1965 1970 1975 1980 1985 1990

In v e s tm e n t, lo g d e v ia tio n

1950 1955 1960 1965 1970 1975 1980 1985 1990

L B ack A

M A G : L o n g - R u n I m p a c ts o f S h o r t - R u n P u b lic I n v e s t m e n t 1 / 4 n d H a o t ia n X ia n g

Full S am ple

O u tp u t p e r c a p ita , lo g d e v ia tio n C o n s u m p tio n p e r c a p ita , lo g d e v ia tio n

1950 1960 1970 1980 1990 2000 2010 1950 1960 1970 1980 1990 2000 2010

In v e s tm e n t p e r c a p ita , lo g d e v ia tio n

0.8 0.6 0.4 0 .2

• 0 .2

1950 1960 1970 1980 1990 2000 2010

L B ack A

M A G : L o n g - R u n I m p a c ts o f S h o r t - R u n P u b lic I n v e s t m e n t 2 / 4 n d H a o t ia n X ia n g

A Lower D epreciatio n Rate

► Can a p la in v a n illa R B C m o d e l w ith a low er d e p re c ia tio n ra te ra tio n a liz e series a ro u n d 1960s? ► S e t th e a n n u a l d e p re c ia tio n ra te o f p u b lic c a p ita l to 5% . Output, log deviation Consumption, log deviation Private investment, log deviation

M A G : L o n g - R u n I m p a c ts o f S h o r t - R u n P u b lic I n v e s t m e n t 3 / 4 n d H a o t ia n X ia n g