Notes 12: Another Example - A Permanent Increase in G

The purpose of these notes is to review the effects of increasing Government Spending on the economy in the short run, long run and really long run – with and without Federal Reserve Action. This is what we did in class this week.

For concreteness, we will assume:

1)The increase in G is not accompanied by an increase in A (the government spending does not affect future TFP in the economy). This is like the increase in spending on homeland security. More airport screeners will not make the U.S. more productive (comparing pre-9/11 to post-9/11).

2)There is no change in taxes accompanying the increase in G.

3)Consumers are PIH. Assume further that they are not liquidity constrained and that they do not follow Ricardian Equivalence. At the end, I may relax some of these assumptions. You should understand how the model works when we have different types of consumers. We spent 1 full lecture on consumer behavior.

4)Assume we that our original equilibrium is at Y*.

A.Short Run Analysis

G increases.

1)First Response: AD shifts out and IS shifts out. AD curve and IS represents the demand side of the economy: Y = C + I + G + NX. If G increases, the demand side of the economy will increase.

P

P1

P0

Y*=Y0 Y1

An increase in G will increase Y on the demand side. Firms will respond by increasing production. How will firms optimally increase production? Firms will only do so if they can raise the price. By raising the price, W/P will fall (nominal wages, W, are fixed in the short run). So, by raising the price of output – holding input prices fixed – firms are more willing to hire workers (real wages will be falling). As a result, N will increase and Y will increase on the supply side of the economy. This is a micro analysis. When input prices are fixed, firms will produce more as prices rise. The reason is W/P will be falling – making it more beneficial to hire more workers. This is EXACTLY what you should have learned in micro as to why supply curves slope up.

Summary: Increase in G will cause increase in aggregate demand (aggregate expenditure). Firms will meet the increased demand by raising prices. The increase in price will make it more beneficial for firms to increase production on the supply side of the economy. W/P will fall, N will increase and Y will increase. Hence, firms will produce more----- this is represented by a move up the aggregate supply curve (NOTE: aggregate supply curve has not shifted in the short run – we just move along the aggregate supply curve!).

The labor market is in disequilibrium. Labor demand has not changed (i.e., A and K have not changed). Nor has labor supply changed (i.e., taxes, population, welfare programs and PVLR have not changed!). The change in real wages – as we will see soon – is temporary. As we know, temporary in changes in wages (by definition) have no effect on PVLR.

In the short run, this is the AD-AS analysis.

What about the IS-LM analysis.

LM1(M0,P1)

rLM0(M0,P0)

r11

r0 z

IS1(G1)

IS0(G0)

Y*=Y0 Y1

As G increases, IS will shift out. But, as we see from AD-AS analysis, P will increase. Holding M fixed at M0, implies an increase in P will cause M/P to fall. As real money supply falls, LM curve will shift to the left.

An increase in IS and a fall in LM will cause r to increase. An increase in r will cause I to fall. THIS IS NOT A SHIFT IN THE IS or AD. This is represented as movements along the curves!!!! You should understand this! Point z on the new IS curve would be the level of Y if the increase in G resulted in no change in interest rates. However, we know that as Y increases, money demand will increase (this is the transaction demand of money). So, as Y increases, Md increases and r increases. As r increases, I will fall. This is a movement along the new IS curve (from point z to point 1). The fall in I due to changes in interest rates is represented as a movement along the IS curve. The AD curve will not shift again as I falls (you should understand this)! This is because the AD curve is just a representation of the IS-LM curve in {P,Y} space. This is already embedded into the AD curve.

So, in the short run, an increase in G will cause the IS curve to shift out. Money demand will increase (as Y increases). r will increase. M/P will fall (as P increases). r will increase further (both money demand and real money supply will be affected). I will fall. We know Y will still increase (that is why interest rates are increasing – you need Y to increase so r will increase – that is the transaction demand of money). The net effect of G increasing and I falling is to have Y increasing!

Let’s summarize formally:

The effects of an increase in G in the short run:

G will increase:

r will increase

I will fall

Y will increase (ΔG > |ΔI|)

W is fixed (by definition of the short run)

P will increase (as AD shifts out)

W/P will fall

N will increase (as W/P falls – move down labor demand curve)

Y will increase on supply side (N increases – this is not surprising since Y=Y).

M/P will fall (as P increases)

C will not change (no change in PVLR).

NOTE: If households were Keynesian, C would also increase. For Keynesian households, Y = a + b Y. As Y increases, C will increase. We would represent this by additional rightward shifts in the AD and the IS curve (because C would also be increasing). I am abstracting from that now. But, you should know that this would occur if households were Keynesian. This would make a good quiz/test question.

NOTE: If households were liquidity constrained, C may also increase. For liquidity constrained households, they may be unable to borrow to smooth consumption. If households have actual C < desired C, they may increase C as Y increases (because they are consuming less than optimal). So, an increase in C would also occur if Y increases. This would cause additional rightward shifts of the IS and AD curve – this would lead to a bigger increase in Y.

NOTE: If households act according to ricardian equivalence, the increase in deficits would cause households to save more (they know they would have to pay the deficit back in the future). In this case, C would fall as G increased. The increase in G would cause the AD and the IS to shift to the right (as above). However, the fall in C (as households save more) would cause the IS and the AD to shift back to the left. According to a strict interpretation of ricardian equivalence, the two effects will exactly offset. In this case, there would be NO effect whatsoever on the economy. P, r, Y all would not change in the short run. As governments save less (G increases), households save more (private savings increases, C falls). National Savings, S, would be unchanged. If S is unchanged, I would be unchanged. If I is unchanged, r is unchanged (changes in r will cause I to change). If r is unchanged, IS curve is unchanged. The increase in G will be exactly offset by a decline in C. This is what ricardian equivalence says (we had a lecture on this AND a quiz question). If households behave as ricardian consumers, there is no effect of deficits on investment (so future generations are unaffected by today’s deficits). The increase in G is exactly offset by the decline in C. Here is how we would represent the AD curve (the IS would be a similar representation).

A permanent increase in G with Ricardian consumers (short run analysis)

P

G increase

P0

C declines

Y*=Y0 = Y1

The increase in G is exactly offset by the decline in C. This is exactly what some people in main stream economics believe. See the Wall Street Journal editorial from Thursday (3/6/03) entitled "Keynes Barks, Ricardo Scratches?"(

Again, most people do not believe in this theory (although the conservative WSJ tends to). Both Glenn Hubbard and Greg Mankiw’s text books teach that Ricardian Equivalence does not hold.

As a side note----you should feel like you are learning a lot! The stuff we are learning in this class is exactly what macroeconomists are talking about. See above WSJ article or Thursday’s NY Times (3/6/03) which had an article “Big Federal Deficits, Bigger Risks” which is by Alan Kreuger(Princeton macroeconomist). In the article he says “According to Ricardian Equivalence, smart consumers would start tucking money away now for future tax increases. But such foresight is uncommon. In a National Bureau of Economic Research paper — titled "What Have We Learned From the Reagan Deficits and Their Disappearance?" — Benjamin Friedman of Harvard documents that personal savings declined as the deficit ballooned in the 1980's and early 1990's. As a consequence, government borrowing crowded out private investment and caused interest rates to rise.”

So, the WSJ believes that the short run effect of deficits look like the picture on this page (i.e., no effect if households are Ricardian). The NY Times believes the short run effects of deficits look like what we drew on pages (1) and (2) of these notes (increase in G causes I to fall). We know understand what the debate is about. The debate centers on how consumers will respond to the increase in deficits (increase in G, holding T fixed). Most economists believe that ricardian equivalence does not hold (at least not completely). So, the consensus among most of us is that what is on pages (1) and (2) of these notes is closer to the truth.

B.Long Run effects of increased G – NO FED INTERVENTION.

What will happen in the long run? If Y > Y*, we know that this is not sustainable. The labor market is temporarily in disequilibrium. Eventually, the labor market will clear. You and I (and all workers) will demand higher wages. As this happens, W will increase. This, by definition, will shift the AS curve to the left. Anything that makes firm production more expensive will shift the AS curve to the left.

The economy will return to Y*. As W increases, P will increase further. However, the end result will be that the %ΔP = %ΔW. How do we know??? Well, when we return to equilibrium in the labor market, our new W/P will equal the original W/P. So, if P increases by 10%, we know W will eventually have to increase by 10%. We will return to the equilibrium real wage in the long run, even though both W and P have changed.

Here is how the AS-AD market will look:

AS1(W1)

P P2

P1

G increases

P0

W increases

Y*=Y0 Y1

So, in the long run, Y will return to Y* (as N returns to N* in the labor market). How do firms respond to the increased demand (AD1)? By increasing prices. The increase in prices will cause M/P to fall even further. As the real money supply falls even further, interest rates will rise even further (from r1 in the short run to r2 in the long run – this is illustrated below). As r increases even more, I will fall even more. Eventually, the decline in I will exactly offset (be crowded out by) the increase in G. In the end, Y will not change on either the demand or the supply side of the equation. On the supply side, N is the same – so Y is the same. On the demand side, G will increase and I will fall. The two changes will exactly cancel each other out. Y (demand side) = Y (supply side). Here is the IS-LM market. The extra increase in P will cause the LM curve to shift in further.

LM2(M0,P2)

LM1(M0,P1)

r2LM0(M0,P0)

r11

r0 z

IS1(G1)

IS0(G0)

Y*=Y0 Y1

As prices rise from P1 to P2 between the short run and the long run, the LM curve will shift in from LM1 to LM2.

Lets summarize the long run impact of increasing G on the economy (given the assumptions made on page 1).

The effects of an increase in G in the short run:

G will increase:

r will increase even more in the long (compared to the short run)

I will fall even more in the long (compared to the short run)

Y will not change between the initial condition and the long run (ΔG = |ΔI|)

W will increase between short run and long run

P will increase even more between the short run and the long run (as AS shifts in)

W/P will not change between the initial condition and the long run

N will not change between the initial condition and the long run

M/P will fall between short run and long run (as P increases by more)

C will not change (no change in PVLR).

Lets stop and test ourselves: Lets look at the following True/False/Uncertain question.

Question 1: Suppose households are ricardian consumers. Suppose G increases permanently and tax rates remain fixed. Suppose the increase in G has no effect on A. Assume, the economy is initially in equilibrium at Y*. Finally, suppose M is fixed through out.

An increase in G will cause inflation in the long run as the economy self-corrects and the labor market clears.

This is a false statement! If households are ricardian consumers, the economy will not be in disequilibrium in the short run. See the analysis on the top of page 4. The increase in G will exactly be offset at by a decline in C. If the economy is not in disequilibrium in the short run, there is no need for the economy to correct itself. The labor market will be in equilibrium in the short run so there will be no reason for nominal wages (W) to increase. If W remains constant, there is no need for the AS curve to shift! As a result, there will be no inflation in the long run.

This statement would be true if households are PIH (non-ricardian) – see above – or, if households are Keynesian!

C.Long run effects of increase in G with Fed intervention!

As G increase, Y will be above Y*. If households are not ricardian consumers, the long run effect of this policy will be increased prices, higher interest rates and lower investment (see analysis above).

So, we know that in the long run, we will be back at Y* with higher prices. Suppose the Fed is interested in managing inflation. Can the Fed undo the inflationary effects of increasing G? The answer is yes.

Note: What we do next is how the Fed behaves. The fed has control over M. As a result, it has control over the AD curve. We saw this last week in class. How can the Fed undo the inflationary aspects of increasing G?

If increasing G shifts the AD curve to the right, the Fed can decrease M which shifts the AD curve to the left. As a result, the AD curve will remain at its original position. Here is how the picture would look.

P

G increase

P0

M declines

Y*=Y0 = Y1

The net effect on the AD curve would be unchanged. In this case, the economy is no long in disequilibrium in the labor market. N will equal N*. Given that we are at the optimal amount of labor in the labor market, there is no need for the economy to correct itself. So, nominal wages in the economy will not change. Hence, there will be no additional pressure on prices in the economy.

What happens in the IS-LM market? As M falls, the LM curve will shift to the left. It can be illustrated as the following:

LM2(M1,P0)

M declines

r2LM1(M0, P0)

r0

IS1(G1)

IS0(G0)

Y*=Y0

The new LM curve will shift in such that Y is restored at Y*. Notice, LM2 in this picture (and the corresponding r2) is EXACTLY the same as the LM curve on the top of page 6. The economy is restored to equilibrium with Fed involvement because M falls. On the top of page 6, the economy is restored to equilibrium because P increases. In both cases, the long run M/P is the same. With Fed involvement, real money supply falls because M falls. With the self correcting mechanism (no Fed involvement), M/P falls because P increases (as workers demand higher nominal wages, W, because they are working so much).

So, in the new equilibrium with Fed involvement it is exactly the same as the long run with no Fed involvement: Y is at Y*, r is higher, I falls, change in G is exactly offset by the decline in I, M/P falls, real wages are at their original level, N is at the original level. The only major difference is that prices do not rise.

The Fed has brought the economy back to Y* by managing the money supply. The Fed, by decreasing M, has shut off any inflationary pressures in the economy. This is exactly how the Fed manages inflation. See on the Web page Greenspan’s testimony to Congress in early 2000. In 1999, he reports raising interest rates because he thought there was inflationary pressures in the economy (Y > Y*). So, he increased i (by decreasing M). He hoped to slow down demand in the economy by slowing down I. This is a shift in of the AD curve.

Our model is a model of the FED!!! What we are learning is exactly the way Greenspan thinks of the world.

Test yourself again.

Question 2. Assume we are initially at Y*. Assume further that households are PIH consumers and do not follow ricardian equivalence. Finally, assume A and tax rates are held constant throughout your analysis.

Suppose G increases permanently. If the Fed brings the economy back to Y*, as opposed to allowing the economy to correct itself, future generations will not be made worse off by the current increase in government deficits.

This is so false! Future generations are hurt by current deficits because current deficits crowd out current investment (as the deficits cause interest rates to rise). If the economy corrects itself, M/P will fall as P increases. If the Fed gets involved, M/P will fall as M falls. In either case, real money supply will fall and r will increase. In either case, I will fall. How much will I fall – in both cases, the fall in I will exactly offset the increase in G (how do we know this???? – because, in the long run, the change in Y will equal 0!!!). So, even if the Fed gets involve, investment will still fall as a result of the deficit. What the Fed involvement does is cause the deficits to be non-inflationary. However, I will still fall today (and future generations will still be made worse off because of the current deficit).

1