Chapter 11
Determining How Much Stuff an Economy Produces
IN THIS CHAPTER
Producing what people want: aggregate supply
Looking at aggregate supply in the long run
Getting to grips with short-run aggregate supply
Recovering economic health
When people are demanding something, someone else needs to fulfill that demand to avoid disappointment all around — whether it’s a baby crying for milk, football fans screaming for a touchdown, or TV viewers lobbying for the return of their favorite cancelled show (Futurama fans are still writing petitions).
Chapter 10 introduces the idea of aggregate demand (the total demand for goods and services in an economy). But clearly, people demanding a certain amount of output aren’t enough — someone needs to supply that output. In a market economy, this important job falls to firms. But companies aren’t robots. They don’t just supply goods because someone wants them. Rather, they have to make decisions about how much to make and what price to charge. What determines this behavior? What determines aggregate supply?
In this chapter you discover what determines how much firms will produce, in the short run and the long run. We also discuss important related topics, including economic growth and price flexibility.
Producing What People Demand: Aggregate Supply
To fulfill the demand from people, firms, and governments, companies have to produce and supply goods and services.
Aggregate supply (AS) is the total amount of goods and services that firms produce in an economy. It’s the other half of the Aggregate Demand–Aggregate Supply (AD–AS) model, which is the central model of the short-run macro economy (see Chapter 10 for details). In the short run, the interaction of aggregate supply and aggregate demand determines how much a country produces (GDP) and how much on average those goods will cost to purchase (the price level), as well as how fast that price level is changing (inflation), plus how much unemployment an economy experiences.
Looking at long-run aggregate supply
Economists distinguish between the short run and the long run because the economy behaves in a different way depending on the time frame you’re looking at. The important thing about the long run is that prices are fully flexible (they can change a lot), whereas in the short run prices are sticky (they don’t change by very much — check out the later section “Pulling apart why prices can be sticky” for some suggested causes of this stickiness).
Saying that prices are sticky is a way of saying something important about the behavior of suppliers. When demand for any product increases, the businesses that make those goods can do one of three things:
- Hold output constant and raise their price until the extra demand is choked off
- Hold the price constant and increase supply enough to meet all the new demand
- Do a bit of both
The same options arise for the workers, who supply labor, and the capitalists who supply capital to the firms. In response to a rising demand for their services, they can supply more labor and capital services, raise their compensation demands, or, again, do some of both.
In the long run, it’s the first option that best describes agents’ behavior in a macro sense. Given enough time, a rise in demand leads to a higher overall aggregate price level P (and higher wages, too) but no more output. In the short run, it’s the second choice that rules. Initially, prices and wages are held in check in response to a demand shock. The result is that, in the short run, a rise in demand evokes mostly an increase in production, whereas a fall in demand induces production to fall. In this section, we review briefly the long-run supply behavior, which is really what the growth model of Chapter 8 is all about. Then we take a look at short-run supply.
Growing output and long-run potential GDP
Remember the growth accounting equation? If you don’t, it’s okay. We’re going to reproduce it here. But you may want to turn to Chapter 8 to get a fuller review.
The growth accounting equation is a way of describing the growth of real GDP or Y over time. It can be written as follows:
gY = (capital’s share) gK + (labor’s share) gL + technical progress
Chapter 8 shows that given steady population growth and technological development, this equation translates into a smooth path with real GDP growing steadily at a rate equal to the population growth rate plus the growth rate of labor productivity. In turn, labor productivity (and hence real wages) grow steadily at a rate that reflects both technical progress and the steadily increasing amount of capital that each generation of workers inherits. So, real GDP grows steadily along the long-run path, too.
(Yes, the growth equation a few lines ago ignores land as an input. Mainly that’s because the supply of land doesn’t grow over time. But even if it did, putting land in the equation doesn’t change much. Because we can get all the insights we need by focusing on just capital and labor, we’ll take this simpler path. Why make trouble?)
The growth process just described gives the actual path of real GDP in the long run. Be careful here, though. We’re not saying that actual GDP will be on the growth path in every single period. What we’re saying is that if you had to predict real GDP for, say, ten years from now, your best guess would be the value projected by the long-run growth path ten years out.
That’s also a way of saying that in the long run, real GDP is equal to the maximum sustainable output that the economy can supply given our supply of inputs and the state of technology. In turn, this means that along the growth path, real GDP expands enough each year to absorb all the net new capital (gK) and all the new workers (gL) and the extra output that these inputs can now produce due to technical progress.
We’ve italicized supply and maximum because they’re essential to a full understanding of what the long-run model is saying. First and foremost, it’s a supply-driven model. Output is totally determined by what the economy can produce given the technology and the supply of capital and labor inputs.
What about demand? What if people don’t want to buy all this output? Good questions. The answer in the long run is that insufficient demand is never a problem. In the long run, households, businesses, the government, and foreign traders always want to purchase what the economy can supply. Why? Because in the long run, prices (and wages) will fall to whatever level is necessary to raise demand to this level. Indeed, the best way to define the long run is as however long it takes for this adjustment to occur. That’s why the price level P doesn’t appear anywhere in the growth accounting equation. Because P will just adjust to whatever level is needed, the supply side of the economy tells us where real GDP will be without any reference to P.
The word maximum is important, too. We could also use the term potential. The long-run path that the growth model traces out is the path of what Chapter 3 calls potential GDP. In the long run, prices and wages adjust to keep the economy realizing its true potential, that is, producing the maximum sustainable amount it can at any time given the amount of capital and labor inputs available as well as the technology known at that time.
By the way, the word sustainable is important, too. For short periods of time the economy might be able to exceed its potential. Workers and machines can be pressed above their normal limits for a while. But this extra output cannot be sustained in the long run.
Looking at long-run aggregate supply … in the short run
Figure 11-1 illustrates a possible path for real GDP over time. As in Chapter 6, we’ve made this a linear path by putting the logarithm of Y on the vertical axis. But this is just a mathy transformation to simplify the picture. It doesn’t change anything fundamentally.
© John Wiley & Sons, Inc.
FIGURE 11-1: The short run is a small Interval of time Δt.
The figure isolates a short interval of time, Δt, between time t1 and time t2. We’re going to treat this time interval as the short-run period. Note that if the interval is not “too” long, the difference between potential GDP at time t1 and time t2 is pretty small. As a result, we can approximate potential GDP over such a short interval as with just one number — say, the average of GDP at the two times. Over a year, for example, potential GDP might grow by about 2.5 percent. So, even if the time between t1 and t2 is two years long, implying 5 percent growth, the average of the two annual values will be just 2.5 percent off from either the value at t1 or the value at t2. This approximation is even more accurate for shorter intervals.
Drawing the LRAS curve
We’ll use Y* to denote the level of aggregate supply that lies on the economy’s long-run growth path for a specific short-run period. Again, Y* is the real GDP that the economy could produce in that short run if it fully employed all its labor and capital inputs — in other words, it’s potential GDP at that time.
As we just said, Y* doesn’t depend on the price level. It is simply a given value of GDP — one that corresponds to the economy being on its long-run trajectory. Therefore, if we graph Y* against price P we simply get a vertical line through Y* showing that Y* is the same no matter what the price level is. Such a graph is shown in Figure 11-2. The long-run aggregate supply, or LRAS, curve is a vertical line through potential or natural GDP Y*.
© John Wiley & Sons, Inc.
FIGURE 11-2: Shifts in aggregate demand change only the price level — output is unchanged.
Figure 11-2 also shows two possible aggregate demand curves: AD1 and AD2. The important thing to note is that if we take Y* as the amount the economy will supply, it doesn’t matter which aggregate demand curve we assume. If supply is truly given at Y*, then real GDP has to be at Y*, too. This just requires that as aggregate demand shifts, say, from AD1 to AD2, the price level falls from P1 to P2.
Another way to say this is that if prices were perfectly flexible even in the short run — if they could quickly move from P1 to P2, — movements in aggregate demand would have no effect on real GDP. They would only alter the price, and the economy would just stay at Y* regardless of where the aggregate demand curve is. If this were the case — if producers could easily raise and lower their price over a short period of time by any amount necessary to keep the economy at full employment — then the short run would be just a snapshot of the long run. Every period, the economy would simply move to the level of GDP consistent with full employment as it progressed along its long-run growth path.
This is the reason we call the vertical line through Y* in Figure 11-2 the long-run aggregate supply (LRAS) curve. It would describe the economy’s short-run supply behavior if prices were perfectly flexible — that is, if the economy behaved in the short run exactly as it does in the long run.
Note that because the economy is at full employment when Y = Y*, measured unemployment will be at the natural rate of unemployment (derived in Chapter 7). Because Y* is the output level that corresponds to unemployment being at the natural rate, macroeconomists sometimes use a shorthand to combine the two concepts by calling Y* the natural rate of output or just natural GDP.
Getting to Grips with Short-Run Aggregate Supply
It’s an exaggeration (but not a grossly inaccurate one) to say that prior to the 1930s most economists who thought about the macro economy tended to believe that price flexibility was the norm. After all, prices on the stock market can rise or fall by several percent in a day and certainly in a week. Why should the prices of goods and/or the price of labor (wages) be any different? We’ll go over some possible answers to that question in a bit. The important thing for now is that starting in about the 1930s, macroeconomists began to change their views about price flexibility and short-run aggregate supply.
Two closely related events led macroeconomists to turn away from their long-held views about the short-run. One was the Great Depression. The other was the 1936 publication of The General Theory of Employment, Interest, and Money by British economist John Maynard Keynes.
Admitting the reality of depression
In late 1929, the U.S. stock market sharply declined. From a high of 381 in early September, the Dow Jones Industrial Average began a long slide to the dramatic events of October 29th, commonly known as “Black Tuesday,” when the market “laid an egg,” as Variety proclaimed, and crashed to 230. Although a number of modest “bull” and “bear” markets followed over the next few years, the general trend was decidedly down. The low point came in mid-1932 when the Dow Jones average closed at just a bit over 41.
The collapse of U.S. stock prices wiped out the wealth of many of the country’s richest families. It also led to a more general financial crisis and banking collapse as more than 9,000 local and regional banks failed. In those days, there was no federal deposit insurance, so those with deposits in these banks — farms, families, and small businesses — lost their money entirely. This financial disaster and the tremendous uncertainty that it generated led to a huge contraction of aggregate demand. If prices were fully flexible, this leftward shift of the AD curve would simply have led to a decline in prices from, say, P1 to P2 in Figure 11-2, and the economy would have remained at Y* and full employment — in other words, we would have remained on the LRAS curve. That didn’t happen.
Real GDP started falling in 1929 and continued to do so for nearly four years into early 1933. In total, it fell by nearly one-third over that time while the unemployment rate climbed to nearly 25 percent. The recovery that followed was not strong enough to return real GDP to potential until 1941. Not surprisingly, the velocity of money (see chapter 5) also plummeted and similarly stayed below its pre-crash level for the rest of the decade.
This is not to say that prices were completely rigid. Indeed, the aggregate price level P fell every year from 1929 to 1933, resulting in a cumulative fall of over 30 percent at the wholesale level. But that decline was not nearly enough to neutralize the fall in demand and keep GDP at its potential or natural rate Y*.
If the drop in GDP had been a lot less severe and if the gap with potential had been eliminated much more rapidly, macroeconomists might have concluded that a short-run model built on flexible prices that continuously kept actual GDP close to potential and the long-run path was a reasonable approximation to reality. Or possibly, if the collapse had been confined to the U.S., the economic debacle might have been written of as a perverse example of “American exceptionalism.”
However, the Depression spread quickly and lasted in many countries for a similarly long time. The evidence of a decade of very high unemployment and sluggish recovery across many different countries could not be ignored. Economists had to face the fact that, for a substantially long short run, the “invisible hand” did not work so well to keep the economy humming along.
Theorizing generally
Keynes’s treatise, The General Theory of Employment, Interest, and Money, was published in 1936, roughly in the middle of the decade of depression. Although Keynes’s ideas had been circulating informally for some time, the book’s publication was a watershed event. It was in many ways Keynes’s book that framed the analysis of aggregate demand that macroeconomists use today. He was the first to argue that fluctuations in aggregate demand were common and that prices were “sticky” in the short run.
As a result, Keynes’s writing challenged the idea that real GDP was determined by aggregate supply even in the short run. Instead, his analysis called everyone’s attention to aggregate demand as the primary determinant of short-run GDP and to which aggregate supply responded more or less passively. And Keynes didn’t stop there. He went on to describe macro policies that the government could use to manage aggregate demand and so counteract the business cycle — to “cure” the economy of its depression.
Thus, Keynes’s work was incredibly timely. At the very moment when the United States and other countries found themselves in a deep depression, Keynes’s book offered both an explanation for it and a prescription for how to get the economy back on track. The Swedish economy appeared to use Keynesian policies successfully to climb out of the Depression relatively quickly. In the U.S., the New Deal policies of the early years of the Roosevelt administration seemed to have a similar effect. Over the past 80 years, there is pretty solid evidence supporting Keynes’ view that wage/price stickiness is real so that a decline in aggregate demand leads to decline in real GDP. His book was a persuasive and lasting contribution.
The Keynesian model did not deny the importance of supply in the long run. Keynes himself was a little agnostic on this point, but later Keynesians have generally agreed that the growth model given in Chapter 7, which assumes full price flexibility, is a good description of the economy’s long-run behavior. It was more that Keynes viewed the short run as what really matters for mortal human beings. Yes, everything might work well in the long run, but, as Keynes famously remarked, “In the long run, we are all dead.”
Supplying more or less in the short run
If wages and prices are sticky, the aggregate supply curve is not vertical in the short run. Instead, a typical short-run aggregate supply curve (SRAS) looks like the one shown in Figure 11-3. This says that in the short-run, prices fall when real GDP falls but not enough to keep the economy at Y*.
© John Wiley & Sons, Inc.
FIGURE 11-3: The SRAS curve implies that shifts in AD change output as well as price level.
Think again of the early Great Depression period. Prices did fall, but so did real GDP. That’s what the SRAS in Figure 11-3 says will happen if aggregate demand falls. As the aggregate demand curve shifts from AD1 to AD2, the price falls along the SRAS from P1 to P2 while output falls from Y* to Y2.
Note that as output falls below Y*, the fact that businesses are producing less means they need fewer workers and machines. Layoffs increase. Factories close or work fewer and shorter shifts. Unemployment rises above the natural rate. Recession ensues. Figure 11-3 is a revealing picture but not a pretty one.
Pulling apart why prices can be sticky
As Keynes saw, the core difference between the short run and the long run was aggregate price stickiness. Keynes was a shrewd observer of the world around him. In 1925, when Great Britain restored the pound to its pre-war gold standard parity, Keynes argued that this led to an overvaluing of the pound that could only be offset by restrictive monetary policy that would force British prices and wages down. (See our discussion of net exports and the real exchange rate in the previous chapter.) In turn, he predicted that this would prolong the high unemployment and sluggish growth that the U.K. had already been experiencing in the first half of the 1920s, exactly as in Figure 11-3.
Events proved Keynes right. The U.K. unemployment rate rose above 8 percent in 1926 and stayed stubbornly at or near 7 percent until the debacle of the Great Depression. Whatever economic theory might say, Keynes felt the empirical evidence was very clear. Wages and prices have built-in inertia, and it’s especially hard to get them to fall.
For Keynes, this kind of evidence from the real world was enough. But formal economic analysis requires more than just informed observation. Observing that prices are sticky in the short run is all very well, but we also want to know why. What causes them to be so? Over time, a number of answers have emerged. Each reveals a separate insight. Together, they provide a fairly robust justification for the view that wages and prices display some short-run rigidity.
The benefits don’t always outweigh the costs
Imagine that you run a small camera shop. At the start of the quarter, you work out the likely demand at different prices and the cost of selling and serving different quantities. After reviewing this data, you then select the best or the profit-maximizing price. Life is good.
But what if your calculations are off a bit? Or what if things are different from what you thought they would be? For example, what if the price you set turns out to be 10 percent higher than the profit-maximizing price given the new, lower level of demand? That is, what if demand is a fair bit lower at the price you chose than you originally predicted?
Obviously, if you had to do it again, you’d set a lower price. But that’s not the issue now. The decision before you is whether to cut your price given that it’s currently too high or keep it where it is. Of course, resetting it involves a little bit of work — and expense. If you have lots of different cameras and accessories, you need to go around updating all the prices in your shop and make corresponding adjustments in your valuation books. It’s not a Herculean task, but it’s not a trivial one either.
Economists call the costs associated with changing prices menu costs. The reason is pretty clear. Restaurants issue printed menus. They don’t incur the expense of reprinting the menus to reflect new prices each and every day as demand for the restaurant’s services goes up and down. They may not even do it each week. Keeping the price constant may not be a bad strategy once we recognize that changing it incurs such menu costs.
Equally important, the benefit or profit gain of getting the true profit- maximizing price is likely to be small for two reasons:
- Although your high price hurts you by limiting the amount you sell below the optimal amount, it is a higher price. You sell less but what you sell goes for a higher fee.
- By selling less you save on costs. You don’t need your employees to work as much overtime. You don’t have to keep the store open as many hours. You can maybe cut back on your inventory.
When these effects are combined, it turns out that having the “wrong” price will cost you some profit but not necessarily a lot. It’s what economists call a second-order effect. It’s there, but it’s not huge. And it may very well not be large enough to cover the menu cost of changing your price.
Relative prices are important
Of course, not all firms have significant menu costs. For example, large online retailers with sophisticated automated systems may have small menu costs. Even so, these firms may still find it optimal to limit how much they cut their price. Why? Because they care about their relative price — their price compared with what their rivals charge. If a firm’s rivals are holding prices constant, then the firm itself doesn’t need to cut its own price very much to pick up some rival customers.
Relative prices are important for another reason, too. Many industries are dominated by a small number of firms. This includes credit cards, aircraft construction, cable television, and domestic mass beer production, to name just a few. For the firms in such industries, the outbreak of a price war is a constant fear. Given that fear, each firm may be very unwilling to cut its price in the face of a drop in demand. Unless it knows that its rivals will understand this action, any firm that lowers its price will be doing so relative to the price of its rivals. It therefore runs the risk of being seen as an aggressor trying to steal customers from others — just the kind of behavior that can anger its rivals and trigger the price war that the firm fears.
Menu costs and relative price concerns mean that even small changes in demand cause significant declines in real GDP. Remember, total dollar spending is nominal income or PY. Suppose this falls by 10 percent. Then the sum of the drop in P and the drop in Y has to be 10 percent, too. Suppose half the firms hold their prices constant while the other half cut their prices by 5 percent to avoid too big of a change relative to the prices of the firms keeping price constant. Then the aggregate price falls by just 2.5 percent, meaning that output Y has to fall by the remaining 7.5 percent. That’s a big drop in real GDP.
Contracts are often written in nominal terms
Many if not most contracts are written using the nominal price (or wage). For example, if a consumer has signed a contract with a firm that promises to provide some service over the year at a certain (nominal) price, she expects that price to be honored even if an increase in AD means that the price should increase. Likewise, the firm expects the consumer to continue paying that price even if she (and all other consumers) decides to buy less than originally expected. Thus, consumers trade the ability to get a low price during periods of slack demand in return for avoiding a high price during periods of peak demand. A similar story could be told about wages.
Nominal wage contracts in which the wage is set by a negotiated labor contract for a year or more seem especially common. These may reflect what economists call insider-outsider considerations. This approach recognizes that it’s the workers inside the firm who negotiate labor contracts, not those outside it who also want to work there. This is important because most drops in aggregate demand are relatively small, and the typical firm may be able to reduce its output enough just by not replacing the normal amount of workers who quit or retire. In that case, it’s workers outside the firm — who would normally be hired to replace those leaving — who suffer unemployment in a downturn.
In this setting, the workers inside the firm may prefer a contract that specifies a specific money wage, for example, $25/hour, while letting the firm determine hiring. To the extent prices are based on costs (wages), such a contract will also implicitly fix prices as well. As a result, the nominal wage contract guarantees both employment and the real wage for those inside the firm, while at the same time rigidifying wages and prices.
Misperceptions about the price level
The following suggested reason for sticky prices is a clever one: Essentially, people confuse an increase in the nominal price of whatever they’re selling for an increase in its real price. This mistake causes them to oversupply the market with goods. Only when they realize that the real price is unchanged do they return to their earlier production.
Say you’re a baker. Every day after baking your loaves of bread, you take them to the marketplace to sell. Unknown to you, aggregate demand has increased, causing the price level to increase. Therefore the price that you can sell your bread for has also increased. You respond by singing “Happy Days Are Here Again” and rushing home to bake as many loaves as you can in order to sell to market.
What you don’t realize, though, is that all prices in the economy have also increased, so the real price you’re receiving for your bread is unchanged. Only the nominal price has increased. At the end of the week, you go to do your weekly shopping and notice that the prices of all goods, including those of the supplies you need, have risen. Only then do you understand that the real price of bread is the same as it was last week. Armed with this understanding, the following week you go back to your usual amount of bread production, this time singing, “Won’t Get Fooled Again.”
The importance of market power
Because recessions and the resulting output gap tend to last a while, the misperceptions story might be a little hard to defend in this era of rapid information and communication. You’ll know pretty quickly that your bread price increase is matched by price increases throughout the economy and adjust your baking pretty quickly then, too.
The other sources of wage and price stickiness may be more robust. Note that all these other explanations invoke some form of market power. Whether it’s firms worrying about the cost of changing their prices, or businesses concerned about rivals’ reactions to any price change, or whether it’s insider workers setting a wage scale, all these stories involve some agents with the power to set a wage or a price. These are not the perfectly competitive players of classical analysis, each of whom has to take the market price or wage as given.
At the base of most modern theories of short-run aggregate supply is a belief that markets are not perfectly competitive. This may be sensible. As noted above, many of the largest industries are populated by a few firms that have some power to set the company’s price. Unlike perfectly competitive markets where prices move up and down readily, the prices in imperfectly competitive markets are set by firms (or workers). Hence, they don’t change until somebody inside those firms decides to change them.
Recovering economic health
As Keynes grudgingly acknowledged, and as we’ve emphasized from the start, macroeconomists regard the long-run growth path as an anchor. The economy may drift away from that path from time to time but: a) it will not drift infinitely far, and b) there are always forces pushing the economy back toward the path. As horrific as the Great Depression or, to a lesser but still deplorable extent, the Great Recession was, GDP did not collapse entirely — far from it. And in these cases and others, the economy has naturally tended to recover, however weakly.
All this is a way of saying that our best guess is, again, that the decline in GDP from Y* to Y2 shown in Figure 11-3 won’t be permanent. Eventually, the economy will return to producing its potential, and real GDP will equal Y*.
How will this happen? That’s the kind of penetrating question we’ve come to expect from you, dear reader. In truth, the answer is a little complicated, but the intuition is straightforward. To see how the full recovery works in practice, consider the demand scenario shown in Figure 11-4.
© John Wiley & Sons, Inc.
FIGURE 11-4: Short run: AD fall reduces prices to P2. Long run: Prices fall to P3, output returns to Y*.
As before, we start from a full equilibrium with output at the potential or natural level Y* (point A). Aggregate demand then falls from AD1 to AD2. If prices were fully flexible, this would leave output unchanged at Y* and cause the price level to fall to P3 (point C). In the short run, however, prices are sticky, and the fall in AD causes output to fall from Y* to Y2 (a movement along the SRAS1 curve from point A to point B). You can also see that, even though prices are sticky in the short run, they aren’t completely fixed. The short-run price level does fall from P1 to P2.
As more time passes, prices in the economy are able to adjust more completely to the drop in demand. This shifts the SRAS curve gradually from SRAS1 to SRAS2. Prices continue to fall as the SRAS2 curve moves outward along the AD2 curve. At point C, the economy has fully recovered.
Note the nature of this process. Officially, a recession is defined as a fall in GDP. This is the movement from point A to B. Subsequently, we enter the recovery phase where real GDP is closing the gap with potential as we move from B to C. During the first part, unemployment rises. During the recovery, unemployment falls, but full employment is not restored — unemployment does not reach the natural rate — until the output gap is closed and Y = Y*.
Two final comments: First, we have told the story for the case of a fall in aggregate demand from AD1 to AD2. However, we could have told essentially the same story in reverse by starting with aggregated demand curve AD2 and then letting it shift out to AD1.
Second, the time from the start of the recession at A to its end at B and then to a full recovery at point C may be measured in years. During this time, potential GDP Y* will be growing (moving to the right). So, actual GDP will have even further to go to close the gap and restore full employment. This is why a complete recovery can take a long time. As of this writing, the U.S. economy has still not quite fully closed the output gap between actual and potential GDP emanating from the Great Recession of 2007–09.