tunes and downright misery of others. But do they?

Suppose a premature frost damages the buds on citrus trees this year, thereby reducing the ftiture harvest. Seeing this, people will naturally expect next years citrus crop to sell for more than this years crop. As a result, incentives are created for withholding some of the previously harvested crop for resale next year at higher prices when the crop failure is felt. Anybody who responds to these incentives is specidat-ing. Although we cannot cover the complex markets which facilitate such speculation, nonetheless, withholding a portion of the crop from the market will assuredly raise the current price of ciuus. On the other hand, selling next year, when the fall impact of the freeze is felt, will assuredly lower the price of citrus compared to what would otherwise occur. As Figure Xni shows, speculation therefore tends to raise the present price and reduce the fiiture price, smoothing the time path of prices.

At die same time, speculators cause citrus to be carried across time to the forthcoming period of increased scarcity, clearly a desirable goal. Indeed, without speculation no incentive to incur agricultural storage costs would exist and nothing would be saved for future use. Consumption would be forced to rise and fall stricdy with the size of the harvest. .As Figure XIV illustrates, speculation tends to reduce the quan-ut>- available in die present while increasing the quandty av-ailable in the future, smoothing the time path of quantit).

Profes.sional speculators, many of whom could scarcely di.stingtush an orange from a tomato, perform another function: they are specialists in risk management. Wliile citrus farmers, wholesale companies, and

even grocery markets can (and do) speculate over citrus prices, many of these traders prefer to minimize their risks by arranging conlracuiallv fixed hiiuio prices through professional speculators. Such arrange-meius allow these producers and middlemen to specialize in what they do relatively best farming, whole-sahng. retailing, etc. while letting professional speculators handle risk, presumably what they do best. .\s we saw in Section I, specialization permits increased output. Consequently, when speculators judge correcily, consumers are made better off

at about when speculators are wrong-Clearly, carrying a crop from a high price period fori resale into a low price environment results in a mi.sal- location of resources: it will make societ)-worse offbvj amplifying price fluctuations. At die same time, ii will decidedly make speculators worse off They will bu\j when the price is high and resell at a loss when it is \o\q While speculators cannot possibh always be right afte all, speculation is risky - speculators who are cousis tendy wrong will be liquidated by the market as the incur rising losses. The market not only generate incentives to store goods for the future, it assures th; in the long run only competent speculators will .surriw to do so.

ELASnCITY OF DEMAND

Although the law of demand tells us that th3 quantity demanded responds inversely to changes i£ price, it tells us nothirig about the magnitude of i response, which can vary enormously from demaii curve to demand curve and even along the sa curve. The standard measure of the demand respon to a change in price is known as the price ela.sticit) i demand (called elasdcit)- of demand for short) and formally defined as:

Percentage Change in Quantit\ Percentage Change in Price

Consider a small price increase of .some percent another along a certain demand curve. lis clear ht the law of demand that the price increase will accompanied by a decrease in the quaiitir. demande again by some percent or anothei. Similarly, a cenrage decrease in the price implies a perceiita increase in the quantity as we reverse directions aIo tlie same demand curve. Thus, in bodi cases we opposite signs in the numerator and dcnomiiiatoij formula (1) above, making a negative numt

We customarily ignore the negative sign of the elasticity of demand. For example, suppose along a certain demand curve that a 1% increase in the price would be accompanied by a 2% decrease in the quantit) demanded. Then we would calculate as follows:

Percentage Change iu Quantit) 2% 2 Percentage Change in Price 1%

Note here that by ignoring the negative sign we define e<i to be a positive rather than a negative number. Moreover, the demand elasticity of 2.0 also implies that a 2% reduction in the quamirs sold along the demand .curve would cause price to rise by 1 %. Similarly, a 3% increase (decrease) in the price would cause a 6% decrease (increase) in the quaiuity sold given our Ed of 1.0 (6%/3% = 2.0).

Since Ed is a measure of the demand respon-fesiveness to a change in price, its convenient to catalogue its possible values. At one extreme, we might [bave no quantity response to a change in price, which ; to say, that the numerator in (1) would be zero. At le other extreme, the quandty purchased could be IBtplosively responshe to even a small change in the price; in this case, ea would approach Infinit). Hence, [js clear that varies between zero and infinity. We jreak these possible \alues into categories in Table 1 fclow.

A demand cur\e with an elasucity everywhere iiial to zero, which is to say, perfecdy inelasdc at all ¹ices is depicted in Figure XV below. Here the term igrfecdy inelasdc" tells us that the quantity demand-I is completely unresponsie to changes in the price, jr example, in Figure XV the same quantity would be ided no matter what the price. Since this polar

case violates the law of demand, wed not expect to see such a demand curve in the real world. Note that a glance at Figure XV tells us at once that = 0 at all prices along this curve. The only other demand curve in which Ed can be determined without a calculation occurs in the polar opposite demand curve of Figure XVI. Here equals infinity, which is to say the curve is perfectly elasdc.

Normally, however, elasdcit) varies with the price along the demand curve. For example, a special type of demand curve often used for convenience is a linear (straight line) one. Figure XVII is typical of such curves and displays the for extremely small price changes in the ricinity of different prices along the curve. Although we shall defer the method of calculation until later, you can see that Ca varies between zero and infinity as we increase price, reaching the value of 1.0 (unity) at the midpoint as we move from right to left along the curve. The point of unit elasticity will become quite important in Section VI dealing with monopoly.

ELASTICITY AND TOTAL REVENUE

Elasticity of demand is intimately bound up with total revenue (TR) received by sellers of a good or - what is the same thing - the total expenditure paid out by the buyers. Reconsider the total revenue formula below:

TR = P X Q

Now imagine in the formula that the price P were to rise by 1 %, but that the quantity Q were to remain constant. From the formula its clear that total revenue (TR) would rise by exactiy 1%. Now lets change the example around a bit and imagine that price remains

Fig. XV

Fig. XVI

Fig. XVII

£d = oo

ed = 1

ed = 0 ,

Perfectly Inelastic Demand

Perfectly Elastic Demand

constant in the formula, but that quantit) falls by 2%. Its equally clear in this case that TR would fall by exactly 2%. So fai, so good. Next lets put both effects together by moving along a certain demand curve. Suppose that when we increase the price by 1%, the quantity demanded falls by 2%. Then were back with our old example of ea = 2.0 (2%/l%), i.e., an elastic part of the demand curve. But its also plain that two opposing effects are working on total revenue. The first effect, the 1% increase in the price, taken alone, tends to raise total revenue, but the other effect, the 2% reduction in the quantity demanded, taken alone, tends to lower it. The final result must therefore depend upon which of these two opposing effects is the larger. Since the 2% reduction in the quantit) demanded reduces total revenue more than the 1% increase in the price raises it, total revenue must in this case fall. Moreover, turning the 2.0 elasticity example around a bit, its likewise clear that a 1% decrease in the price will end up raising total revenue because now the revenue increasing effect of the 2% quanut) increase will overpower the opposing effect of the 1 % price decrease. Similarly, anyume Ed > an increase in the price lowers total revenue, while a decrease in price raises it.

On the other hand, when the percentage change in price exceeds the percentage change in quantity, i.e., when < 1. total revenue (TR) will rise with a price increase. For example, if price rises by 1%, and quantity demanded falls by 1/2% along a demand curve, then Ed = V2 (V2%/1%). an inelastic value. In this case, the relatively greater percentage increase in price, 1%, will overcome the opposing effect of the 1/2% decrease in quantit), so that total rev-

enue will rise. Now reverse directions and convincej yourself that total revenue will fall if the price were lo fall when Ed < 1-

Finalh", when the percentage change in quan-1 tit) ju.st equals the percentage change in price, i.e.,! when Ed = I, the opposing effects exactly cancel so thatJ total revenue (TR) will be constant with either a price] decrease or increase.

We can summarize these basic results as folloAVs: j

(1) If demand is elastic (Ed > l), the price and] total received by sellers (the total expenditurej of the buyers) change in opposite directions.

(2) If demand is inelasdc (Ed < l)i price and] total revenue (total expenditure) change in the saincj direction.

(3) If total re\enue (total expenditure) is conl stant as price changes, demand is unit elastic (£d =

In Figure XVIII below, we see that total re\v eiiue has decreased from area + to area + fol-J lowing a price increase from Pi to P2, so that deinandj is ela.stic hi this range (Ed > l). A price increase of equal size in Figure XIX, however, results in an] increase in total re\enue from area B + to B + \\ consequently, we know that demand is inelastic in this] range (Ed < 1).

One special case in which Ed = i at all prices] occurs ill Figure XX where total revenue is constant afj all prices. Such a situation might occur when, fon example, UCSB is given a fixed budget for, say, hiringi professors. If the University will spend its entire biid-j get no matter what the gohig salary (price) of profes- sors, the toul revenue received by UCSB professorsi would be constant, so that must in this case equal] unit\.