Consumption and Investment. Graduate Macroeconomics I ECON Cunningham

Transcription

1 Consumpion and Invesmen Graduae Macroeconomics I ECON Cunningham

2 Keynesian Theory Recall ha Keynes argues ha C= C 0 +cy, wih C 0 > 0 and he average propensiy o consume (APC = C/Y) is greaer han he marginal propensiy o consume (MPC = c): C/Y = (C 0 + cy)/y > c, or (1) APC > MPC (2) Moreover, he APC should no be a consan if C 0 is no zero. If C 0 = 0, hen he consumpion funcion reduces o he absolue income hypohesis consumpion is proporionae o income which is no consisen wih Keynes. 2

3 Empirical Verificaion? Keynes followers esimaed he consumpion funcion for he U.S. using he daa from : C = Yd C 0 =26.5 billion > 0 APC > MPC Increases in consumer spending seemed o be less han increases in disposable income, supporing MPC < 1. 3

4 Kuznes Consumpion Daa Kuznes, Simon. Uses of Naional Income in Peace and War, Occasional Paper 6. NY: NBER, Time series esimaes of consumpion and naional income Overlapping decades , 5 year seps Each esimae is a decade average Kuznes, Simon. Naional Produc Since NY: NBER, Exended daa backward o

5 Kuznes Sudy (1) Assumpions: Personal axes and ransfer paymens are small (in his period) Therefore, i is reasonable o use oal income (GNP) as a proxy for disposable income. If a relaionship beween consumpion and disposable income exiss, here should also be a relaionship beween consumpion and GNP. 5

6 Kuznes Sudy (2) Resuls: (1946 sudy) Beween , real income expanded o seven (7) imes is 1869 level ($9.3 billion o $69 billion) Bu he average propensiy o consume ranged beween and Tha is, APC did no vary significanly in he face of vasly expanding income. Problem! 6

7 Kuznes Sudy (3) Years Y C C/Y

8 Second Failure Predicions of pos-wwii period are grossly wrong Keynesian Theory argues ha he average propensiy o save (APS) rises wih income (S = S 0 +sy). Higher pos-war incomes should imply excess saving. Excess saving is more han can be absorbed by invesmen. Therefore he excess saving will resul over-invesmen or hoarding, and herefore in unemploymen. Will we go sraigh back o he Depression? Comparison of he forecass wih he acual resuls sugges ha: consumpion was under -prediced saving was over -prediced IMPLICATION: major deerminans in he behavioral equaions mus be missing! 8

9 9

10 Reconciliaion wih Keynes Theory? (1) Arhur Smihies, Economerica, Uses per capia Yd and C, and a ime rend. Argues ha he Cons. Funcion is really Keynesian, bu jus looks Kuznesian because of he shifs in he funcion. C * Kuznes = 1925 = 1924 = 1923 Says he daa poins jus happen o line up o fi Kuznes consumpion funcion. Y * 10

11 Smihies Resul Smihies esimaes yielded he equaion: C * = Y * (-1922) Or equivalenly, C * = [ (-1922)] Y * Which is of he form C = C 0 + cy, he equaion of a line. 11

12 Reconciliaion (2) Reasons for shifs: Migraion of people from farms o ciies (mus buy goods) Shif in disribuion oward greaer equaliy (poorer save less) Rise in he perceived sandard of living (luxuries become necessiies) For hese reasons each agen (per capia) should increase his or her consumpion. 12

13 Modigliani ges involved Modigliani, Franco. Flucuaions in he saving-income raion: a problem in economic forecasing, in Sudies in Income and Wealh vol. 11, Conference on Research in Income and Wealh. NY: NBER, 1949, pp When Franco Modigliani (1949) esimaes Smihies relaion over a differen ime period, he analysis compleely breaks down. Bu, Modigliani is hooked... 13

14 More Problems 14

15 15

16 Habi Persisence Theory Noe: Duesenberry and Modigliani boh presened similar resuls a he Economeric Sociey Meeing of Duesenberry (1947) noed ha in 1935 dissaving grew as a percenage of income. Dissaving was greaer in1935 han in he relaively prosperous year Why? Households mus sacrifice saving o defend (aemp o mainain) heir sandard of living. Duesenberry and Modigliani can reconcile he shor-run and long-run consumpion funcions, bu canno explain he negaive relaionship beweem curren income and consumpion ha someimes occurs. 16

17 Duesenberry s Habi Persisence Duesenberry assumes ha he consumers defend heir highes level of consumpion. Thus: S Y = Y F Y peak For no oher reason han simpliciy, he assumes lineariy. S Y = a + Y b Y peak 17

18 18 More More Duesenberry Duesenberry Noe: Y C Y S and Y C Y S so C S Y = + = + = 1 1 = peak Y Y b a Y C ) (1 So ha peak Y Y Y b a C = 1 If Y grows seadily, hen ) (1,, ) (1, 1 bk a wih Y C or Y bk a C and k Y Y Y Y peak = = = = = λ λ Kuzne s Kuzne s Resul Resul

19 Dusesenberry s Rache Effec C C Y 19

20 Brown s Habi Persisence Brown. Habi Persisence and Lags in Consumer Behavior, Economerica 20 (July 1952). Brown ries: C = a 0 + a Y 1 + a 2 C peak and C = a 0 + a Y 1 + a 2 C 1 He finds he second specificaion slighly superior and more easily consisen wih he downward shifs. (Economeric ess.) 20

21 Hamburger and Wealh Theories Hamburger (1951, 1954, 1955) Argues an alernaive explanaion for slugglishness ies consumer response formally o ineremporal uiliy maximizaion. Household ies curren consumpion o a lifeime plan of consumpion: u = u ( c, c +, c + 2,..., c + 1 i Wealh arises from propery wealh and human wealh (invesmens and income). Households consume less under an esae moive. Explains habi persisence rigorously. ) 21

22 Life Cycle Hypohesis (LCH) Franco Modigliani, Alber Ando, and Richard Bloomberg Assumes ha each represenaive agen will die, and knows: when he/she will die, how many periods T he/she will live, and How much his/her life-ime income will be. The consumer smoohs consumpion expendiure over his/her life, spending 1/T of his/her life-ime income each period. 22

23 LCH (2) The consumpion funcion implied by his logic is: 1 C = + T [ 1 1e Y + ( N 1) Y A ] wih he aggregae esimable consumpion funcion look like his: 1 1e C = by 1 + b2y + b3 A 23

24 Income and Consumpion LCH deah Consumpion Saving Income Dissaving T 24

25 Tesing he LCH If he funcion form looks like his: 1 1e C = by 1 + b2y + b3 Ando and Modigliani argue ha expeced fuure labor income is proporional o curren income, so ha he funcion can be reduced o: 1 C = ( b1 + b2β ) Y + b3 When hey esimae his funcion, hey ge: 1 + A A C = 0.72Y 0. 06A 25

26 Criicisms of LCH The households, a all imes, have a definie, conscious vision of: The family s fuure size and composiion, including he life expecancy of each member, The enire lifeime profile of he labor income of each member afer he applicable axes, The presen and fuure exen and erms of any credi available, and The fuure emergencies, opporuniies, and social pressures which migh affec is consumpion spending. I does no ake ino accoun liquidiy consrains. 26

27 Policy Implicaions of LCH Changes in curren income have a srong effec on curren consumpion ONLY if hey affec expeced lifeime income. In Q2 1975, a one-ime ax rebae of $8 billion was paid ou o axpayers o simulae AD. The rebae had lile effec. Maybe George W. hadn heard abou his? The only way here can be a significan effec is if here is a srong liquidiy consrain operaing. This has implicaions for moneary policy. 27

28 Permanen Income Hypohesis (PIH) Milon Friedman (A Theory of he Consumpion Funcion. Princeon Univ. Press, 1957) Assumpions: Perfec cerainy abou: Fuure receips Fuure ineres raes Fuure prices, ec. People save o reduce flucuaions in expendiures People are immoral (or leave bequess) Individual s uiliy funcion: u = u(c,c 1 ) where c is curren period consumpion and c 1 is nex period consumpion. 28

29 PIH, Coninued Assume posiive bu diminishing reurns and posiive cross parials. Also assume ha u is homogeneous of degree one. The budge consrain is: 1 1 c + c1 = q + q = 1+ r 1+ r w e Where q,q 1 are planned real receips from human and nonhuman wealh in each period, and w e is expeced wealh. Opimizaion yields: c = i c 1 = A i ( r ) w B e ( r ) w e 29

30 PIH, Coninued This implies ha consumpion depends upon income only if he income affecs expeced wealh. Therefore, changes in income need no affec curren consumpion. We may view y p = rw e as permanen income wha really affecs he consumpion decision. c = c P + c T y = y P + y T 30

31 PIH, Coninued Where is has permanen (p) and ransiory (T) componens. For each individual agen: c P = k(r,v,x)y P Where r = rae of reurn v = he raio of human o nonhuman wealh (permanen income) x = demographics By summing across individuals (aggregaing), we have: C P = κy P, ha permanen consumpion is proporional o permanen income. 31

32 PIH, Coninued When reduced o a esimable form, his becomes: c = κa 0 Y + λc -1 + θ Which looks remarkably like he Brown (1952) resul. The resul can explain he unusual changes in MPC due o changes in income: If income changes, individuals end o decrease saving (or dissave) o mainain a more uniform spending paern over heir lives. 32

33 PIH (2) Individuals updae adapively heir esimaes of permanen income based on changes in curren income. Tha is, hey learn. Y p p = Y p ( ), 1 + j Y Y 1 0 < j < 1 The resul is ha changes o curren income have lile effec on curren consumpion unless he individual believes ha he changes has long-erm consequences. 33

34 Invesmen Spending Invesmen is he change in he capial sock I n, = K K -1 = Ne Invesmen I g, = K K -1 δk = Gross Invesmen (δk is depreciaion) I = I(r,E) = I(r) Wha abou he expecaions erm? 34

35 The Acceleraor Model (1) Aemps o capure some measure of curren business condiions (growh of he economy or lack of i), and use ha o explain he level of invesmen. 35

36 Acceleraor Model (2) The desired capial sock is proporional o he level of oupu: d K = αy Invesmen is he process of moving from he curren level of capial o a desired level: d I K K n, = 1 We assume ha whaever he capial sock ended up being las period was he level of capial ha businesses acually waned: K d 1 = K 1 = αy 1 36

37 Acceleraor Model (3) This allows us o rewrie: d In, = K K 1 As d I K K = αy αy I n, = 1 1 = α( Y Y 1) n, = α Y Thus invesmen is relaed o he rae of change in oupu. If he economy is growing rapidly, hen invesmen grows rapidly. If he economy is no growing, hen invesmen slows, and ne invesmen (afer depreciaion) may acually be negaive. 37

38 Acceleraor Model (4) As a resul of adjusmen coss and pracical ime-o-build consideraions, he enire adjusmen o he desired capial sock may no be done in one period. The firms may only finance a parial adjusmen. Le λ be he fracion of he gap beween he desired and acual capial sock ha he firms pursue. This leads o: Or, equivalenly, I n d, = λ( K K 1), 0 < λ < 1. I λ( αy K 1) n, = This is referred o as he flexible acceleraor model of invesmen. 38

39 Cos of Capial Approach MEC all over again? Recall ha Keynes argued ha business decision makers compare he expeced revenue sream from he new capial o he cos of capial. The user cos of capial is he oal cos o he firm of employing an addiional uni of capial for one period. The new capial migh be funded by borrowing, selling sock shares, reained earnings, ec. 39

40 Cos of Capial Approach (2) This suggess an invesmen funcion of he form: I I Y, CC, K ) n, = ( 1 If a firm invesed is reained earnings or monies raised by selling sock shares, i could earn he curren ineres rae. So his mus be he opporuniy cos we are looking for. Invesmen mus be relaed o his ineres rae specifically, o he real ineres rae φ, where: φ = r p& e 40

41 Cos of Capial Approach (3) Par of he user cos of capial is he depreciaion rae δ. e So CC = φ + δ = r p& + δ. Bu some governmen programs provide subsidies o firms for purchasing capial. For example, he gov may offer invesmen ax credis. If he porion paid by governmen is τ, hen he effecive cos of capial o he firm is: CC e = ( 1 τ )( r p& + δ ). This leads o an invesmen funcion of he form: e I I Y, r, p&, τ, K ) n, = ( 1 41