There were 63 department stores, 855 grocery stores, 793
motels and hotels, and 2,452 eating and drinking establishments
in the area in 1972 [28, 29]. These statistics were utilized to
develop the aggregate demand curves for the area.13
 The aggregate demand curves for average sizes of establish-
ments over a range in price of $0.40 to $3.20 per thousand gal-
lons are shown in Figure 5. At a price of $0.80 per thousand all
department stores in the Miami SMSA would utilize 108 million
gallons of water per year (Figure 5). Grocery stores would pur-
chase 391 million gallons at this same price. The eating-drinking
establishments and the hotels-motels groups would use consider-
ably greater quantities of water, with hotels and motels using
1,368 million gallons per year. Eating and drinking establishments
used even more at 1,515 million gallons per year (Figure 5).14
 Elasticities of demand vary among the various types of com-
 mercial establishments at any particular price. At a price of
 $0.80 per thousand gallons elasticity estimates range from 0.09
 for the hotels-motels group to 0.86 for the department stores
 group. Elasticity estimates were 0.58 for the grocery stores and
 0.22 for the eating and drinking establishments.1' These differ-
 ences are significant to policymakers and others concerned with
 setting prices for commercial water users. All of the elasticity
 estimates are low when r=$0.80. An increase in the price of the
 water by 10 percent would cause the department store group to
 reduce purchases by 8.6 percent, the grocery group by 5.8 per-
 cent, eating and drinking establishments by 2.2 percent, and the
 motel-hotel group by 0.9 percent, indicating the latter group has
 the most inelastic demand for water."
 The aggregate commercial demand is illustrated in Figure 6.
 This curve represents the total quantity of water that would be
 purchased by department stores, grocery stores, eating-drinking
 establishments, and hotels-motels given the number of estab-

 13These were the latest statistics available. See Appendix B, Table 5
 for information regarding numbers of other types of commercial estab-
 lishments.
 14The data tables showing the actual quantities estimated at various
 prices are in Appendix B, Tables 6-10.
 '1As discussed earlier in this report, the price coefficient on the eating
 and drinking establishment model was not statistically sound. It is ex-
 pected, however, demand elasticity would be very low for eating and
 drinking establishments.
 "'This general statement must be interpreted with caution. These are
 point elasticities, meaning the response at a particular price. If price was
 increased substantially (such as to $1.20 per thousand), the percentage
 change in quantity purchased would be slightly more than predicted.