Unemployment, Inflation, and Economic Policy

Transcription

1 Unemployment, Inflation, and Economic Policy Topic 7 1

2 Goals of Topic 7 Definition of Short-Run and Long-Run Equilibrium. An analysis of demand and supply shocks. Unemployment and Inflation Dynamics: Phillips Curve. Fed s Objectives and Behavior. 2

3 One Caveat on the Shape of the SRAS Curve The Shape of the SRAS Curve is believed not to be linear: The Shape of the SRAS Curve looks more like: P SRAS(A 0,W 0 ) P 0 It is bounded at some technological constraints. It flattens out at low levels of output. 3

4 What is Equilibrium for the Economy? Short-Run Equilibrium: AD = SRAS and IS = LM The Labor Market need not be in equilibrium (nominal wages are sticky, remember). We need not be at the potential level of GDP Long-Run Equilibrium: AD = SRAS = * and IS = LM = * and N d = N s = N* In the long run, by definition, we will move to *. In the long run, by definition, the labor market will clear. 4

5 Graphical Representation of the Economy in Recession * = f(n*,k,a) P SRAS = f(input prices) AD = f(g,pvlr,taxes, f,m,π e ) Below Potential! e sr * 5

6 More on Equilibrium Equilibrium is a point of attraction for the economy: Most macroeconomists believe that, in the absence of shocks, the economy would reach equilibrium after perhaps 5 years. Thus the economy is in equilibrium in the long run (after 5 years). Is the economy ever in long-run equilibrium? Given that shocks are always hitting, the economy is not likely to be in long-run equilibrium at any point in time. et the force of attraction of equilibrium keeps the economy hovering around the equilibrium. Why is the long-run equilibrium point attractive? Because there the labor market clears. Away from * workers are not on their labor supply curve (and firms may be off their labor demand curve). Maximizing behavior by workers and firms push the economy towards long-run equilibrium. Shocks push the economy away temporarily. 6

7 How do we represent recessions? Unofficial Definition - 2 consecutive quarters of negative real GDP growth Our Model: < * We define the output gap as * - e sr Remember: * trends upward overtime (even though in our model - we have it constant - population growth, K grows (we abstract from these trends in our model), TFP is growing every period). It makes things easier to keep * constant initially. Our Model removes the trends in N and K due to population growth and K (by assumption). Questions for the rest of class: What causes recessions? Why would the fed be concerned about inflation when e sr >*? How do we get out of recessions once we are in them? Can the economy correct itself? 7

8 Where Do Recessions Come From? 1. Temporary or Permanent changes in technology (or productivity) - i.e. productivity shocks, drought? 2. Shocks to Expectations (individual beliefs about the future) i.e. stocks are overvalued, a recession will be coming (self-fulfilling prophesies), etc. See Bonus Material on Expectations for more. 3. Increases in relative prices of inputs (like oil prices) 4. Bad (or good) Economic Policy (monetary or fiscal) 5. External Crises (for small open economies). 8

9 Analysis of the Business Cycle: Examples We now start analyzing shocks to the economy that move it away from full-employment output. In the examples that follows such shocks are permanent. ou can also study their temporary counterparts. Convention: the initial shocked variable (say, G, M s, consumer confidence,etc.) has a 1 subscript (example if I tell you we are increasing G then we move from G 0 to G 1 > G 0 ). 9

10 Some Examples 1974 Recession: OPEC - rapid increase in oil prices 1980 Recession: OPEC II Iran Revolution 1982 Recession: Good Fed Policy! (Volker s Recession) 1990 Recession: Consumer Confidence s Japan: End to speculative bubbles, bad policy, debt overhung and liquidity traps Recession: Consumer Confidence (Burst of Stock Market bubble - see reading on the Economist The Kiss of Life? ) Recession: Credit Crunch and Home Prices Collapse 10

11 Example 1: Changes in Consumer Confidence An change in expectations about the future (whether founded or not) can have dramatic effects on the economy. Consumer Confidence Irrational Exuberance Paradox of Savings (an increase in the savings rate will decrease aggregate demand and may as well decrease savings and investment) Households expect the future to be good or bad. No effect on labor demand (A hasn t really changed). Some effect on labor supply (Households expect PVLR to have changed). Households work more or less (depending on whether they think PVLR has decreased or increased). Consumption changes and the AD and IS curves shift Example: Recession (a fall in consumer confidence from the Gulf War) Note: I will often assume small (or no) income effects on labor supply. 11

12 Consumer Confidence:

13 Analyzing a Decrease in Consumer Confidence Example 1: Recession due to Gulf War I. Note: No Effect on A or A f P P 0 SRAS(W 0 ) Assume we start at * Assume consumers are standard PIH (i.e., non-liquidity constrained, non-ricardian). Assume no income effect on labor supply. r AD(C 0 ) LM(P 0 ) W/P N s W 0 /P 0 IS(C 0 ) 13 N* 0 N d

14 Analyzing a Decrease in Cons. Conf. (SR Eq.) P P 0 SRAS(W 0 ) Assume we start at * Assume consumers are standard PIH (i.e., non-liquidity constrained, non-ricardian). Assume no income effect on labor supply. P 1 r e sr AD(C 0 ) AD(C 1 ) LM(P 0 ) LM(P 1 ) W/P Consumer confidence per se does not affect labor supply or labor demand. However in the short-run firms meet the lower demand at lower prices P 1 but higher labor costs W 0 /P 1 N s W 0 /P 1 1 W 0 /P 0 0 IS(C 0 ) e sr IS(C 1 ) 14 N* 0 N d N sr e

15 Analyzing a Decrease in Cons. Conf. (LR Eq.) P P 0 P 1 P 2 r e sr 1 e sr 0 2 AD(C 1 ) SRAS(W 0 ) SRAS(W 2 ) AD(C 0 ) LM(P 0 ) W/P LM(P 1 ) LM(P 2 ) W 0 /P 1 1 IS(C 1 ) IS(C 0 ) 15 W 0 /P 0 Eventually, nominal wages will fall to W 1 (self-correcting mechanism). Nominal wages are fixed in the short run (that got us to point (1)). In the long run, they can adjust. We get back to the labor market equilibrium (0). The fall in nominal wages makes production cheaper which will shift out the SRAS curve. When production is cheaper, firms want to produce more at every given price! 0 = W 2 /P 2 N* 0 N s N d

16 Irrational Exuberance: Increase in Cons. Conf. Mistaken belief in higher A f shifts IS/AD rightward, and shifts * leftward due to lower N* (income effect - may be small!). The boost in perceived PVLR and future MPK raises demand for C and I goods, shifting the IS/AD rightward. The higher perceived PVLR reduces N* and shifts the LRAS. Note: current A does not rise (so labor demand does not shift). In the SR, the economy moves to the new IS/ LM intersection, with higher and higher r. Firms respond to the higher goods demand by producing more (to achieve this they hire more N). The economy moves to > *, N > N*, and U < U*. Prices Increase!!! Irrational Exuberance could cause inflation (higher prices - in both the short run and the long run!) 16

17 Analyzing the Curves The curves have meaning!!! Notice - the slopes of the curves give us different quantitative results. Steepness of curves matter! Look at SRAS-AD (flat (instantaneous) SRAS - versus upward sloping SRAS). The text book considers flat SRAS. In class increasing SRAS Look at the IS-LM (money market vs IS-LM; Fed Interest Rate Targeting). 17

18 Example 2: Supply Side of the Economy - Oil Responsible (partially) for the 1975 and Economic Recessions (OPEC I and OPEC II) Also important for today s economy! Stellar oil prices. Supply shocks cause unemployment and prices to move in the same direction (VER DIFFERNET FROM DEMAND SHOCKS). Supply Shocks - oil and technology! When an input is more expensive, I need to employ less of it and produce less. Demand Shocks: M, G, T and consumer confidence. As we will see soon - supply shocks could have demand effects! 18

19 Analyzing an Increase in Oil Prices Example 2: 1974 (and 1979) as OPEC countries set a world oil embargo. P SRAS(Oil 0,W 0 ) Assume we start at * Assume the production function is not affected. P 0 r AD(C 0 ) LM(P 0 ) IS(C 0 ) 19

20 Analyzing an Increase in Oil Prices: (SR Eq.) P P 1 P 0 r SRAS(Oil 1,W 0 ) 1 e sr 0 SRAS(Oil 0,W 0 ) AD(C 0 ) LM(P 1 ) LM(P 0 ) The economy is initially in equilibrium at point (0). As the AS curve shifts in (due to higher oil prices raising the cost of firm production), prices should RISE and the equilibrium level of output should fall to ( 1 GDP in the short run). Also the AD shifts in because consumption and investment decrease. The short-run equilibrium is at (1) for the economy. We refer to this situation of rising prices (inflation) and high unemployment (low output) as stagflation. 1 0 e sr 20 IS(I 0,C 0 ) IS(I 1,C 1 )

21 Analyzing an Increase in Oil Prices: (LR Eq.) P P 1 P 0 r e sr 1 e sr SRAS(Oil 1,W 0 ) 1 2 * 2 2 * SRAS(Oil 0,W 0 ) SRAS(Oil 1,W 1 ) AD(C 0 ) LM(P 1 ) LM(P 0 ) IS(I 0,C 0 ) IS(I 1,C 1 ) The economy is in LR equilibrium at point (2). Notice that the permanent increase in oil prices diminishes productivity of factors of production. This can be considered like a negative TFP shock. Hence the LR potential output level will be lower, at * 2. What happens in the labor market in short run? Well if < *2, N < N* 2 and people are working less than their optimal. In the short run, nominal wages are fixed. Prices go up. Real wages will fall. What happens in the long run? We will go to * 2 (by definition of the long run, that is what always happens). We will assume we get there via the selfcorrecting mechanism. As people work less then their optimum, firms will want to cut nominal wages. As nominal wages fall, SRAS will shift to the right, restoring general equilibrium at *2. Prices will drop further and the LM will shift back out.

22 Example 3: 1990s and increase in A Increase * (shift out labor demand and shift in labor supply - regardless if N increases or decreases, * will likely increase - the TFP effect will usually dominate). SRAS will shift out (Cost of Production Falls) AD and IS will shift out (C and I will increase, PVLR and MPK will increase) How far will AD and SRAS shift out? Depends - many cases. Summary, will increase in both short and long run. The effect on Prices is ambiguous: they could rise, fall or stay the same, interest rates will likely rise (although, it is not guaranteed if there is a sufficiently large drop in the price level), C will rise in short and long run, I will likely rise in short run and may rise in long run, Tax Revenues will increase, Public saving will rise, Private Saving is uncertain - may stay the same, may rise or may fall - depending on the timing of the technology changes and the instruments used to hold wealth (ie, pensions). 22

23 Analyzing an Increase in TFP - Labor Market W/P W 0 /P 0 0 N 0 2 LR N 2 N 1 1 SR N s N s N d N d An increase in A will increase N d directly. An increase in technology (like an Internet revolution) will increase the marginal product of labor which will make labor more productive (and hence worth more to the firm). And because of change in PVLR as real wages increase, N s will fall. If the substitution effect dominates, * will increase (both A and N rise). If the income effect dominates, it is uncertain whether * increases (A increases and lower N causes to fall). 23

24 Analyzing an Increase in TFP (A) Example 3: The internet Revolution. Late 1990s in the U.S. SRAS(A 0,W 0 ) P P 0 r AD(C 0, I 0 ) LM(P 0 ) IS(C 0,I 0 ) 24

25 Analyzing an Increase in TFP (SR Eq.) P P 0 r * 2 1 * 2 1 SRAS(A 0,W 0 ) AD(C 1, I 1 ) AD(C 0, I 0 ) 1 LM(P 0 ) IS(C 1,I 1 ) 1 SRAS(A 1,W 0 ) IS(C 0,I 0 ) There will be effects on both AS and AD. Since PVLR increased permanently, consumption will rise. This will shift out the AD curve. Additionally, as A increases, the MPK will increase causing firms to invest more. This will increase I and also shift out the AD curve. How far will it shift out? It depends. There is no guarantee that it will shift all the way out to the new potential level of output. We will assume that it only goes part of the way (in reality, it could actually surpass the new *). What about the AS curve? An increase in A will shift out the AS curve (the new technology will make production cheaper so firms will increase production at any given price).

26 Analyzing an Increase in TFP (SR Eq.) This representation of the economy is the one that the Fed thought that we were in during Spring 2000 (read FOMC statements at the time): 1) Technology had improved. 2) Consumption had increased 3) Investment had been rising. 4) Unemployment was lower than the estimated natural rate of unemployment. 5) Prices had been relatively stable DESPITE the rapid GDP growth. 6) Fed thought the labor market was tight another word for wages possibly rising in the future. 26

27 Analyzing an Increase in TFP (LR Eq.) P P 2 P 0 r * 2 1 SRAS(A 0,W 0 ) SRAS(A 1,W 1 ) 2 SRAS(A 1,W 0 ) 2 AD(C 1, I 1 ) AD(C 0, I 0 ) LM(P 2 ) LM(P 0 ) IS(C 1,I 1 ) In the Long Run Nominal wages would rise to clear the labor market (workers are working so much this can only be a temporary effect. In order to keep workers in the labor market, firms will have to raise wages (actually, the workers will demand it)). As nominal wages increase, the SRAS curve will shift in restoring equilibrium at the new long-run output 2 *. As the SRAS curve shifts in, what happens to prices? Prices will rise! The increase in prices is EXACTL what the Fed was worried about! IS(C 0,I 0 ) 27 * 2

28 Reviewing The Data: From First Class: We saw that some falls in GDP were associated with no increase in prices. We also so that some falls in GDP were associated with large increase in prices. Do our theories reconcile these facts? ES! Demand shocks result in recessions without a corresponding rise in prices. ES! Supply shocks (like changes in oil prices) result in recessions with a large corresponding positive change in prices. ou should really understand the difference between demand shocks (things that primarily affect AD) and supply shocks (things that primarily affect AS) on the economy - their implications are much different from each other. 28

29 Economic Policy Monetary and Fiscal Policy: May they help in stabilizing the economy from demand and supply shocks? NOT AN EXACT SCIENCE - How much stimulus is necessary to move the economy to *. Policy Creates uncertainty as economic agents try to anticipate Fed/Government rules. Some argue avoid using no stabilization policy (just a simple Quantity Theory representation) or suggest using very simple rules. Policy often entails long and variable lags! Some research says that every sustained period of large inflation is due to the Fed! 29

30 Fiscal Policy in Action: Stabilizing Output Another stabilization mechanism is to shock AD through a change in G (fiscal policy). Not necessarily the best approach (we know that an increase in G crowds out private investment by decreasing savings and increasing r i.e. investment moves up along the I d curve in the goods market). It suffers from even longer lags than monetary policy due to the budgeting process. It may be inflationary. Graphical Analysis of an increase in G follows. 30

31 Analyzing an Increase in Gov t Spending (G) P P 0 SRAS(W 0 ) Assume we start at * Assume consumers are standard PIH (i.e., non-liquidity constrained, non-ricardian). Assume no income effect on labor supply. r AD(C 0 ) LM(P 0 ) IS(C 0 ) 31

32 Analyzing an Increase in G P P 1 P 0 SRAS(W 0 ) AD(G 1 ) Assume we start at * Assume consumers are standard PIH (i.e., non-liquidity constrained, non-ricardian). Assume no income effect on labor supply. Wages are sticky. r AD(G 0 ) LM(P 0 ) IS(G 1 ) IS(G 0 ) e sr 32

33 Analyzing an Increase in G (SR Eq.) P SRAS(W 0 ) Prices increase to P 1 in the Short-run equilibrium. P 1 P 0 AD(G 1 ) r 1 AD(G 0 ) LM(P 0 ) LM(M 0, P 1 ) 0 e sr 33 IS(G 1 ) IS(C IS(G 0 )

34 Analyzing an Increase in G (LR Eq.) P P 2 P 1 P SRAS(W 1 ) SRAS(W 0 ) 1 AD(G 1 ) Nominal wages rise to offset the increase in prices. At any price, a rise in the nominal wage also rises the real wage, so firms employ less labor and produce less. r 2 1 AD(G 0 ) LM(P 0 ) LM(M 0, P 2 ) 0 IS(G 1 ) LM(M 0, P 1 ) IS(C IS(G 0 ) e sr 34

35 Analyzing an Increase in the Money Supply P P 0 SRAS(W 0 ) Assume we start at * Assume consumers are standard PIH (i.e., non-liquidity constrained, non-ricardian). Assume no income effect on labor supply. r AD(C 0 ) LM(M 0,P 0 ) IS(C 0 ) 35

36 Analyzing an Increase in the M s (SR Eq.) P P 1 P SRAS(W 0 ) AD(M 1 ) Assume we start at * Assume consumers are standard PIH (i.e., non-liquidity constrained, non-ricardian). Assume no income effect on labor supply. r AD(M 0 ) LM(M 0,P 0 ) LM(M 1, P 1 ) 0 1 IS(C 0 ) e sr 36

37 Analyzing an Increase in the M s (LR Eq.) P P 2 P 1 P SRAS(W 1 ) SRAS(W 0 ) 1 AD(M 1 ) Nominal wages rise to offset the increase in prices. At any price, a rise in the nominal wage also rises the real wage, so firms employ less labor and produce less. In the long-run M 0 /P 0 = M 1 /P 2 r AD(M 0 ) LM(M 0,P 0 ) LM(M 1, P 1 ) LM(M 1, P 2 ) IS(C 0 ) e sr 37

38 Some Thoughts on Money Let us look at the Fed increasing the nominal money supply We are going to look at money supply setting (as opposed to interest rate setting the Fed can only set either one or the other L d matters!). Assume we start at *. Fed increases M s and LM shifts right, r falls and I increases. AD shifts out - increases ( > *). Puts upward pressure on nominal wages (notice the key role of nominal wages as an indicator of future inflation). As nominal wages rise - SRAS shifts in (up). Prices rise and we return to * (basically - this is what is predicted by the quantity theory of money!!!!) 38

39 An Example of Monetary Policy: Stabilization Consumer Confidence Falls By a Lot P SRAS(W 0 ) Same drop in consumer confidence as we saw in Example 1. P 0 P 1 r e sr AD(C 0 ) AD(C 1 ) LM(P 0 ) LM(M 0,P 1 ) As we saw the economy will eventually correct itself and bring us back to *0. The drawback is that the selfcorrecting mechanism works slowly when < (it takes time to cut nominal wages). Let s for a moment think about how the Fed could bring us back to. IS(C 0 ) e sr 39 IS(C 1 )

40 An Example of Monetary Policy: Stabilization The Fed increases money supply to stabilize the economy. P P 0 P 1 r e sr e sr 40 AD(C 1 ) SRAS(W 0 ) AD(C 0 ) LM(P 0 ) LM(M 0,P 1 ) LM(M 1,P 0 ) IS(C 1 ) IS(C 0 ) As M increases, I increases and the AD curve shifts back out. Notice (this is subtle) when I increased between the short run and the long run when the economy corrected itself, the AD curve did not shift! The reason? I increased as P fell and M/P increased. The response of changing P on Investment (and as a result ) is why the AD curve slopes down. In the case of the selfcorrecting mechanism, we just move along the AD curve when the economy corrects itself (because of the price effect on M/P). When M changes the AD curve shifts. It says that at every given P, higher M is higher M/P resulting in lower interest rates and higher investment and higher. That means, at every given P an increase in M leads to higher ( a rightward shift in the AD curve).

41 LR Eq. : Compare Without and With Stabilization Fed does not stabilize: longer & lower P Fed stabilizes: faster & higher P P SRAS(W P 0 ) SRAS(W 0 ) SRAS(W 1 ) P 0 P 1 P 2 r 0 1 e sr 0 2 AD(C 0 ) AD(C 1 ) LM(P 0 ) LM(M 0,P 1 ) LM(M 0,P 2 ) P 0 P 1 r 02 1 e sr 0 AD(C 0 ) AD(C 1 ) LM(P 0 ) LM(M 0,P 1 ) LM(M 1,P 0 ) 1 2 IS(C 0 ) 1 2 IS(C 0 ) e sr IS(C 1 ) 41 e sr IS(C 1 )

42 An Example of Monetary Policy: Inflation Control Consumer Confidence Increases By A lot (Irrational Exuberance) P 0 P SRAS (W 0 ) r AD(C AD(C 0 ) 1 ) LM(P 0 ) IS(C 1 ) IS(C 0 ) 42

43 Goals of the Fed Fed s primary goals are to foster economic growth and job creation and restrain inflationary pressure. The Fed wants to set r so that r = r*, = *, U = U*, N = N* ( these are equivalent). = * (0-2% inflation). The Fed wants to raise r when r < r*, > *, U < U*, N > N* > * The Fed wants to lower r when the opposite conditions hold. 43

44 Fed Timing Recession Begins First Fed nominal rate cut December months later November months later July months later July months later March months before December months before 44

45 Rules vs. Discretion Should a central bank have a specific policy rule: Rules are explicit and may be mandated by law Money should grow at 4% per year (Friedman preferred rule). Explicit inflation target: UK has an inflation target of 2% (+/- 1%) See main web page Rules may be implicit and known by all economic agents The Fed will target the inflation rate at 2-4% per year. 45

46 Benefits of Rules Commits Central Bank to Some Policy Removes Inflation Temptation Creates a more stable economic situation: Individuals and Firms can anticipate the central bank actions. No surprises! Prevents central bank from thinking too much - the economy is so complex that Fed policy can have delayed impact and is usually initiated too late! Central Bank actions can often be destabilizing. (Freidman, Lucas: Both prefer simple rules). 46

47 Benefits of Discretion Central Bank uses all information possible to make the best decision at the time. The Fed uses a discretionary rule. The members of the bank vote on a monetary policy eat each FOMC meeting. The policy is not dictated by some explicit rule. Benefits of Discretion: Allows Central Bank to choose from competing policy goals. (Sometimes inflation targeting is not best for the economy) The Fundamentals of the Economy may change/or evolve over time - the rule becomes outdated. How do we know where the economy is relative to * and target inflation rate? Information flows slowly and is complex to analyze. 47

48 Hawks and Doves How does the Fed balance price stability ( = *) and full employment (U = U*) when they conflict? [For example, when > * and U > U* at the same time.] Hawks put more weight on * (and have lower values for it). U.K., Canada, New Zealand. Bundesbank before; European Central Bank now? Doves put more weight on staying near U* (and have higher *). ellen tends to puts the same weight on each. Some of my research relates to communication of those weights to the market. 48

49 Fed s Policy Behavior: The Taylor Rule Q: Can we approximate Fed s decisions in a simple, synthetic way? John Taylor of Stanford University: the Fed s behavior under Alan Greenspan (August 1987 to 2006) and Janet ellen (today) are welldescribed by: Taylor Rule: i = r* + π e +.5*(π e -π e *) +.5*( - *)/* i = the nominal federal funds rate. r* = the real fed funds rate target (this is the r consistent with =*). π e = expected inflation. π e * = target inflation. = real GDP. * = equilibrium real GDP (-*)/* is the output gap or GDP gap. A positive output gap means overheating and potentially rising inflation (labor markets will demand higher wages). Taylor used r* = 2% and π e * = 2%. The Taylor Rule explains about 2/3 of quarterly variation in the fed funds rate since 1987 (a lot). 49

50 Notes on the Taylor Rule Fed economists find an even better fit with 1 on the GDP gap term. This does NOT mean that ellen uses this rule, it is just that Fed behavior looks very similar to this rule. Furthermore, the Fed tends to smooth interest rates relative to the Rule: This quarter s actual i =.6*(Last quarter s actual i) +.4*(Taylor Rule s i) Studies have found that other G7 Central Banks (e.g., the Bundesbank, ECB) have also followed versions of a smoothed Taylor Rule. Read the speeches by ellen, Bernanke, Greenspan and Gramlich to see the Fed s take on such subject. 50

51 A Look At Liquidity Traps Nominal Interest Rates Are Bounded At Zero! People Believe that there will be deflation in the future! Real Interest Rates are Large and Positive. Fed Would like to cut rates (to stimulate (shift out AD) which will put upward pressure on prices), but nominal rates cannot go below zero! The Fed is helpless. How do they stimulate investment and consumption when they cannot cut nominal interest rates? This describes the situation in Japan during the late 1990s. Japan has experienced deflation AND nominal rates are close to zero. Central Bank of Japan is helpless. Same in US for

52 r 0 r 0 = - e r A Look At Liquidity Traps 1. Investment-Saving I 0 2. Liquidity-Money S d I d I, S The nominal interest rate paid by nonmonetary assets NM is bound from below at 0 and the real interest rate paid is bound from below at - e. The reason is that i = r + e so if r < - e then r + e < 0 and hence i < 0. But at that point M (the monetary assets which pays i= 0) would become suddenly more rewarding that NM (i.e. nobody would hold NM). This implies that trying to push real rates further down through increases in M (hence shifting LM right) is not possible beyond the i = 0 bound. IS-LM Equilibrium e - e LM UNAVAILABLE L d IS M 0 /P 0 52

53 Demand Side Effects of Deflation Deflation can make borrowers - either consumers or firms, worse off. As we saw early in the course, unexpected inflation makes borrowers better off. They expected to pay a certain real rate and when inflation is higher and the nominal rate is fixed, the real rate they pay is lower (in terms of lost purchasing power). If the economy experiences unexpected deflation, the opposite happens: borrowers are paying more in terms of lost real purchasing power when there is unexpected deflation. Borrowers, both consumers and firms, will essentially be poorer. (Even though, there is another side of the market - somebody s got to lend to them, this could still have large effects on consumption and investment). This demand side effect of deflation is called debt overhang or debt deflation. <<Note, even the government is paying higher than expected real rates on their debt>>. Even if the deflation is expected, large transfers can occur from borrowers to lenders because nominal interest rates are bounded by zero (shuts down lending channels). 53

54 Change in Prices versus Inflation Labor Markets are forward looking! If the purchasing power of workers nominal wages is systematically eroded by unexpected changes in prices, workers will try not to lose systematically. Change in prices become dynamic. Two Examples: Cost-Push Inflation: Consider an increase in production costs. Accommodating supply shocks can lead to persistent inflation (the Fed in the mid and late 70s). Workers see this, and adjust. They ask for higher nominal wages in anticipation, this shifts up the SRAS, and this pushes prices up Demand-Pull Inflation: Policy makers try to permanently keep the economy above its potential level by stimulating demand. Wages will keep adjusting then (you can t fool all the people all the time). 54

55 Graphing Accommodation of Inflation Cost-Push and Accommodation P AS(Oil 1 ) AS(Oil 0 ) P 1 P 0 * AD(M 0 ) 55

56 Graphing Accommodation of Inflation Cost-Push and Accommodation P P 2 AS(W 1,Oil 1 ) AS(W 0,Oil 1 ) AS(W 0,Oil 0 ) P 2 P 1 P 0 AD(M 1 ) * AD(M 0 ) 56

57 How do we get out of Cost-Push Inflation? Fed Can Break the Inflation! Reset, expected inflation rates. The Volker Recession (Paul Volker, an extremely underrated Fed Chair). Through a cold-turkey cut of money supply and with a substantial effort to modify the market s perception of the Fed he managed to cut inflation from double digits to 4%. It did cost the economy a short but deep recession though. 57

58 Graphing Demand-Pull Inflation Graph of a sustained policy to keep above *. P SRAS(W 0 ) P 1 P 0 AD(M 1 ) * AD(M 0 ) 58

59 Graphing Demand-Pull Inflation Graph of a sustained policy to keep above *. May induce an upward spiral in P. P SRAS(W 1 ) P 2 SRAS(W 0 ) P 1 AD(M 2 ) P 0 AD(M 1 ) * AD(M 0 ) 59

60 Business Cycle and Inflation Let us now stop and think. What can we say about the relationship between output and prices in light of our analysis of shocks and the possible stabilizations through policy? We know that in recessions labor supply is higher than labor demand (disequilibrium in the labor market), hence we have unemployment. In booms labor supply is lower than labor demand (disequilibrium in the labor market again), hence we have an utilization rate of labor that is higher or some story for being able to extract more labor (see discussion of efficiency wages in the textbook ch. 11). We also know that prices change. Positively after a positive demand shock such as an increase in money supply or an expansionary fiscal policy is implemented (increase in G). Can we see a pattern between unemployment and inflation? 60

61 The Phillips Curve Discoverer: British economist A.W. Phillips. Discovery: A negative correlation between the unemployment rate and the inflation rate across years within a country in the 1950s. Corr(π, u ) < 0 The correlation was also negative in the U.S. and other countries through the 1960s. Old Keynesians in the 1960s: We have found a stable, exploitable tradeoff between the rate of inflation and the rate of unemployment. We can permanently lower the rate of unemployment at the cost of a permanently higher inflation rate. No. The negative relation is only between unexpected inflation and cyclical unemployment. Expectations-augmented Phillips Curve (Friedman-Phelps) Phillips Curve: π = π e -h*(u-u) u is the natural unemployment rate (structural and frictional unemployment only). 61

62 Unemployment and Inflation

63 Friedman and Phelps: Expectations Matter ou cannot exploit the trade off between inflation and unemployment unless you constantly surprise people. That is, if workers and firms are rational they will incorporate the economic policy that exploits the trade off in their decisions and will not allow the government to fool them systematically by increasing M s or G. Milton Friedman in 1968: Theory predicts that the only AD shocks that are going to The LR Phillips Curve is vertical. Vindicating evidence: The Phillips Curve broke down after 1970 s. Over time in the U.S., higher money growth just leads to more inflation and no higher real GDP. Across countries, higher money growth just leads to more inflation and no higher real GDP. Real GDP actually appears to be hindered by high levels of inflation. 63

64 Unemployment and Inflation

65 Unemployment and Inflation

66 This image cannot currently be displayed. Why is Long-Run Phillips Curve Vertical? Self-Correcting Mechanism ou Cannot Sustain > * forever Quantity Theory π LR Phillips Curve u SR Phillips Curve u 66

67 Is There a Short-Run Trade Off? Short-Run tradeoff between the unemployment rate and inflation rate changes: Inflation tends to fall in years following U > U*. The cost of a permanently lower inflation rate is a temporarily higher unemployment rate. Inflation tends to rise in years following U < U*. The cost of temporarily lowering the unemployment rate is a permanently higher inflation rate. This is why Fed is currently worried about inflation! Demand Shocks cause a negative relationship between P and U (early 80s)! Supply Shocks cause a positive relationship between P and U (the 70s and the late 90s)! 67

68 Conclusions We have experimented with different types of shock to our aggregate economy. Supply and demand shocks. We can describe in a simple, intuitive way most past booms and recessions. Also our simple graphical analysis is helpful for describing possible future scenarios. We now have a way of modeling Output Stabilization Policies and Inflation Control Policies. The Fed s Objectives and the Taylor Rule. Defined the Phillips Curve and the Expectations augmented Phillips curve. 68