1 00:00:00,040 --> 00:00:02,460 The following content is provided under a Creative 2 00:00:02,460 --> 00:00:03,870 Commons license. 3 00:00:03,870 --> 00:00:06,910 Your support will help MIT OpenCourseWare continue to 4 00:00:06,910 --> 00:00:10,560 offer high quality educational resources for free. 5 00:00:10,560 --> 00:00:13,460 To make a donation or view additional materials from 6 00:00:13,460 --> 00:00:16,180 hundreds of MIT courses, visit mitopencourseware@ocw.mit.edu. 7 00:00:21,900 --> 00:00:25,950 PROFESSOR: So today what we're going to do is continue our 8 00:00:25,950 --> 00:00:28,620 discussion of supply and demand. 9 00:00:28,620 --> 00:00:31,430 This is sort of introduction week, if you will. 10 00:00:31,430 --> 00:00:33,450 We've kind of talked about supply and demand, and you 11 00:00:33,450 --> 00:00:35,310 guys, rightly, immediately were on to where do those 12 00:00:35,310 --> 00:00:35,960 curves come from. 13 00:00:35,960 --> 00:00:37,730 And that's what we'll start next week. 14 00:00:37,730 --> 00:00:41,320 But what I want to do today is talk some more about what 15 00:00:41,320 --> 00:00:44,390 determines the shapes of supply and demand curves and 16 00:00:44,390 --> 00:00:47,400 just think about an overview of how we think about supply 17 00:00:47,400 --> 00:00:50,310 and demand interacting in a market and what determines how 18 00:00:50,310 --> 00:00:54,720 responsive individuals and firms are to prices. 19 00:00:57,330 --> 00:00:58,805 And, once again, remember everyone should have a handout 20 00:00:58,805 --> 00:01:00,530 that you should have picked up in the back on your way in. 21 00:01:00,530 --> 00:01:02,380 So everyone should have a handout. 22 00:01:02,380 --> 00:01:05,930 What we talked about last time was the sort of qualitative 23 00:01:05,930 --> 00:01:08,320 effects, the qualitative version of the supply and 24 00:01:08,320 --> 00:01:09,360 demand model. 25 00:01:09,360 --> 00:01:11,770 We talked about what happens when a supply curve shifts, 26 00:01:11,770 --> 00:01:13,590 what happens when a demand curve shifts. 27 00:01:13,590 --> 00:01:16,860 We talked about how either a supply shock or a demand shock 28 00:01:16,860 --> 00:01:18,190 could lead to the price being increased. 29 00:01:18,190 --> 00:01:19,420 But they could have very different effects on 30 00:01:19,420 --> 00:01:21,660 quantity, et cetera. 31 00:01:21,660 --> 00:01:24,200 What we didn't talk about is how big these effects are. 32 00:01:24,200 --> 00:01:25,230 I made up some numbers. 33 00:01:25,230 --> 00:01:26,380 I threw them on the graphs. 34 00:01:26,380 --> 00:01:28,610 But I didn't talk about where the size of those 35 00:01:28,610 --> 00:01:30,050 effects come from. 36 00:01:30,050 --> 00:01:33,490 And where they come from is the shapes of the supply and 37 00:01:33,490 --> 00:01:34,180 demand curve. 38 00:01:34,180 --> 00:01:36,700 And that's what we'll talk about today is what determines 39 00:01:36,700 --> 00:01:40,500 the shapes of supply and demand curves. 40 00:01:40,500 --> 00:01:42,790 And that will be the focus of today's lecture. 41 00:01:42,790 --> 00:01:46,220 I'll talk both theoretically about what determines these 42 00:01:46,220 --> 00:01:49,390 shapes and empirically about how economists go about 43 00:01:49,390 --> 00:01:52,990 figuring out the shapes of supply and demand curves. 44 00:01:52,990 --> 00:01:58,370 So, to think about this, let's start with Figure 3-1, which 45 00:01:58,370 --> 00:02:02,290 is a standard market diagram we had last time. 46 00:02:02,290 --> 00:02:06,160 With an initial equilibrium at point E1, with an initial 47 00:02:06,160 --> 00:02:08,990 price P1 and a quantity Q1. 48 00:02:08,990 --> 00:02:10,600 That's the equilibrium that's stable. 49 00:02:10,600 --> 00:02:16,280 Because at that price P1, consumers demand Q1 units, and 50 00:02:16,280 --> 00:02:19,430 suppliers are willing to provide Q1 units. 51 00:02:19,430 --> 00:02:21,620 So that's a stable equilibrium. 52 00:02:21,620 --> 00:02:23,040 Now we have some supply shift. 53 00:02:23,040 --> 00:02:24,830 Last time we talked about somehow a 54 00:02:24,830 --> 00:02:26,080 pork-specific drought. 55 00:02:28,310 --> 00:02:32,120 That leads the supply to shift inward. 56 00:02:32,120 --> 00:02:35,600 So the supply curve rises to S2. 57 00:02:35,600 --> 00:02:39,920 At that new price, initially, you would have excess demand. 58 00:02:39,920 --> 00:02:42,610 But quickly the price increases to shut off that 59 00:02:42,610 --> 00:02:43,750 excess demand. 60 00:02:43,750 --> 00:02:45,620 And you end up with a new equilibrium with a higher 61 00:02:45,620 --> 00:02:48,950 price, P2, and a lower quantity Q2, and new 62 00:02:48,950 --> 00:02:50,930 equilibrium point E2. 63 00:02:50,930 --> 00:02:51,460 OK? 64 00:02:51,460 --> 00:02:53,340 And we talked that through last time. 65 00:02:53,340 --> 00:02:56,360 What I want to talk about this time is well, what determines 66 00:02:56,360 --> 00:03:00,260 the size of that shift from Q1 to Q2 and that price increase 67 00:03:00,260 --> 00:03:02,040 from P1 to P2? 68 00:03:02,040 --> 00:03:11,600 What's going to determine it is the elasticity of supply 69 00:03:11,600 --> 00:03:13,160 and demand. 70 00:03:13,160 --> 00:03:17,580 The elasticity of supply and demand is how much do supply 71 00:03:17,580 --> 00:03:18,790 and demand respond? 72 00:03:18,790 --> 00:03:22,500 Do the quantities supplied and the quantities demanded 73 00:03:22,500 --> 00:03:25,860 respond when the price changes? 74 00:03:25,860 --> 00:03:29,970 When we say, how elastic is demand, what we mean is how 75 00:03:29,970 --> 00:03:33,320 sensitive to price is the quantity demanded. 76 00:03:33,320 --> 00:03:39,190 Or, alternatively, what is the slope of that demand curve? 77 00:03:39,190 --> 00:03:43,050 So the slope of the demand curve will be the sensitivity 78 00:03:43,050 --> 00:03:46,290 of quantity demanded to the price consumers face. 79 00:03:46,290 --> 00:03:51,040 And that will determine the market responsiveness. 80 00:03:51,040 --> 00:03:53,350 In economics, it's always true that the best way to think 81 00:03:53,350 --> 00:03:54,970 about things is to go to extremes. 82 00:03:54,970 --> 00:03:56,800 You have to remember that extremes don't exist in the 83 00:03:56,800 --> 00:03:58,740 real world. 84 00:03:58,740 --> 00:04:01,070 But it's a useful teaching device 85 00:04:01,070 --> 00:04:01,740 to think about extremes. 86 00:04:01,740 --> 00:04:05,260 So let's think about one extreme case in Figure 3-2. 87 00:04:05,260 --> 00:04:10,380 Let's think about the case of perfectly inelastic demand. 88 00:04:10,380 --> 00:04:13,090 Perfectly inelastic demand, that's where there's no 89 00:04:13,090 --> 00:04:14,150 elasticity of demand. 90 00:04:14,150 --> 00:04:18,120 What that means is that demand for a good is unchanged 91 00:04:18,120 --> 00:04:20,720 regardless of the price. 92 00:04:20,720 --> 00:04:23,880 So perfectly inelastic demand is a case where demand for the 93 00:04:23,880 --> 00:04:26,480 good is unchanged regardless of the price. 94 00:04:26,480 --> 00:04:28,810 That would lead you to have a vertical demand curve at a 95 00:04:28,810 --> 00:04:29,880 given quantity. 96 00:04:29,880 --> 00:04:33,250 What this says is regardless of the price, people always 97 00:04:33,250 --> 00:04:36,430 demand Q. 98 00:04:36,430 --> 00:04:40,820 Can anyone tell me what would cause demand to 99 00:04:40,820 --> 00:04:42,120 be perfectly inelastic? 100 00:04:42,120 --> 00:04:45,620 In what types of situations would demand be-- 101 00:04:45,620 --> 00:04:46,770 it's never perfectly inelastic-- 102 00:04:46,770 --> 00:04:49,030 would demand be relatively inelastic? 103 00:04:49,030 --> 00:04:49,754 Yeah? 104 00:04:49,754 --> 00:04:51,450 AUDIENCE: [INAUDIBLE PHRASE]. 105 00:04:51,450 --> 00:04:53,310 PROFESSOR: It's all about substitutes. 106 00:04:53,310 --> 00:04:57,160 When there's no substitutes, when there's nowhere to go, it 107 00:04:57,160 --> 00:04:59,600 doesn't matter what the price is. 108 00:04:59,600 --> 00:05:02,470 When there's no substitutes, demand will be perfectly 109 00:05:02,470 --> 00:05:06,930 inelastic, because you have to have Q. It doesn't matter what 110 00:05:06,930 --> 00:05:07,900 the price is. 111 00:05:07,900 --> 00:05:12,180 Because there's no substitute for that good. 112 00:05:12,180 --> 00:05:14,860 So if you wanted amount Q of that good for any reason, 113 00:05:14,860 --> 00:05:16,080 you're always going to want that amount Q 114 00:05:16,080 --> 00:05:17,590 no matter the price. 115 00:05:17,590 --> 00:05:19,650 So a perfectly inelastic good would have no substitutes. 116 00:05:19,650 --> 00:05:21,810 So you'd always want Q no matter what. 117 00:05:21,810 --> 00:05:23,050 Can anyone think of an example? 118 00:05:23,050 --> 00:05:24,750 There's no perfectly inelastic good in the world. 119 00:05:24,750 --> 00:05:25,930 But what sorts of goods? 120 00:05:25,930 --> 00:05:26,412 Yeah? 121 00:05:26,412 --> 00:05:28,340 AUDIENCE: Medicines. 122 00:05:28,340 --> 00:05:29,230 PROFESSOR: Medicines. 123 00:05:29,230 --> 00:05:31,590 Now, not necessarily all medicines. 124 00:05:31,590 --> 00:05:34,140 So give me an example of a medicine which would be more 125 00:05:34,140 --> 00:05:35,390 or less inelastic. 126 00:05:39,850 --> 00:05:42,630 So I don't even need a medical name. 127 00:05:42,630 --> 00:05:44,030 What sort of treatments? 128 00:05:44,030 --> 00:05:46,130 AUDIENCE: Like heart attack maybe? 129 00:05:46,130 --> 00:05:48,440 PROFESSOR: Yeah, something which is sort of lifesaving. 130 00:05:48,440 --> 00:05:51,220 The best thing that we often use is insulin for diabetics. 131 00:05:51,220 --> 00:05:53,250 Diabetics without getting that insulin to manage their 132 00:05:53,250 --> 00:05:55,100 diabetes will die. 133 00:05:55,100 --> 00:05:56,606 That seems like that's something where there's not a 134 00:05:56,606 --> 00:05:57,360 whole lot of substitutes. 135 00:05:57,360 --> 00:05:59,890 The substitute is dying. 136 00:05:59,890 --> 00:06:01,300 So basically that's where demand 137 00:06:01,300 --> 00:06:02,850 is relatively inelastic. 138 00:06:02,850 --> 00:06:05,400 Or a heart transplant, when you get a heart transplant or 139 00:06:05,400 --> 00:06:08,400 any kind of transplant, you have medicine you take so you 140 00:06:08,400 --> 00:06:10,390 don't reject the transplanted organ. 141 00:06:10,390 --> 00:06:13,770 That sort of medicine demand should be very inelastic. 142 00:06:13,770 --> 00:06:15,280 Elastic drug, well, our favorite 143 00:06:15,280 --> 00:06:17,540 example is always Viagra. 144 00:06:17,540 --> 00:06:20,570 It's something where you'd think that you can probably 145 00:06:20,570 --> 00:06:21,820 survive without it. 146 00:06:24,220 --> 00:06:26,350 And people would want less Viagra if you charged a lot 147 00:06:26,350 --> 00:06:28,720 more for it than if you charged less for it. 148 00:06:28,720 --> 00:06:33,010 So elasticity is going to be about substitutability. 149 00:06:33,010 --> 00:06:35,640 And that's going to determine inelastic demand. 150 00:06:35,640 --> 00:06:38,720 Now, what happens with inelastic demand when there's 151 00:06:38,720 --> 00:06:40,710 a supply shock? 152 00:06:40,710 --> 00:06:43,570 When supply increases, what happens? 153 00:06:43,570 --> 00:06:47,910 Well, in that case, there can never be excess demand, 154 00:06:47,910 --> 00:06:50,040 because demand doesn't change. 155 00:06:50,040 --> 00:06:54,280 So all that happens is price just increases. 156 00:06:54,280 --> 00:06:58,050 If there's inelastic demand, and there's a supply shock, 157 00:06:58,050 --> 00:07:01,630 then all that happens is an increase in price and no 158 00:07:01,630 --> 00:07:05,070 change in quantity. 159 00:07:05,070 --> 00:07:07,610 So with inelastic demand, quantity doesn't change for a 160 00:07:07,610 --> 00:07:08,280 price increase. 161 00:07:08,280 --> 00:07:09,540 Price just goes up. 162 00:07:09,540 --> 00:07:11,990 From a supply shock, prices just goes up. 163 00:07:11,990 --> 00:07:14,710 Now, let's consider the opposite. 164 00:07:14,710 --> 00:07:16,970 Let's look at Figure 3-3 and think about 165 00:07:16,970 --> 00:07:18,420 perfectly elastic demand. 166 00:07:21,870 --> 00:07:27,390 Perfectly elastic demand is demand where consumers, 167 00:07:27,390 --> 00:07:29,730 essentially, don't care about the quantity. 168 00:07:29,730 --> 00:07:31,440 They just care about the price. 169 00:07:31,440 --> 00:07:34,805 That is, there are infinitely good substitutes. 170 00:07:34,805 --> 00:07:37,615 A perfectly elastically demanded good would be one 171 00:07:37,615 --> 00:07:40,080 where there are, essentially, perfect substitutes. 172 00:07:40,080 --> 00:07:42,060 An inelastic good is where there's no substitute. 173 00:07:42,060 --> 00:07:44,640 A perfectly elastic good would be where there's perfect 174 00:07:44,640 --> 00:07:47,200 substitutes. 175 00:07:47,200 --> 00:07:49,980 Technically, if a good is perfectly elastically 176 00:07:49,980 --> 00:07:54,250 demanded, then you are completely indifferent between 177 00:07:54,250 --> 00:07:56,550 that good and a substitute. 178 00:07:56,550 --> 00:08:01,400 Well, if you're completely indifferent, then if the price 179 00:08:01,400 --> 00:08:05,760 changed at all, you would immediately switch. 180 00:08:05,760 --> 00:08:08,660 And so the price can't change. 181 00:08:08,660 --> 00:08:10,080 What's an example? 182 00:08:10,080 --> 00:08:13,550 Once again, there's no good example of a 183 00:08:13,550 --> 00:08:14,420 perfectly elastic good. 184 00:08:14,420 --> 00:08:15,273 Yeah? 185 00:08:15,273 --> 00:08:16,100 AUDIENCE: Candy. 186 00:08:16,100 --> 00:08:16,790 PROFESSOR: What? 187 00:08:16,790 --> 00:08:17,140 AUDIENCE: Candy. 188 00:08:17,140 --> 00:08:17,990 PROFESSOR: Candy. 189 00:08:17,990 --> 00:08:18,500 OK. 190 00:08:18,500 --> 00:08:21,280 So you've got your Wrigley's gum. 191 00:08:21,280 --> 00:08:22,930 I like the sugar-free, minty gum. 192 00:08:22,930 --> 00:08:25,275 You've got Orbit and Eclipse. 193 00:08:25,275 --> 00:08:27,060 And I go to the store, and they're all pretty much the 194 00:08:27,060 --> 00:08:27,590 same price. 195 00:08:27,590 --> 00:08:30,210 If Orbit was more than Eclipse, I just buy Eclipse. 196 00:08:30,210 --> 00:08:30,660 They're the same. 197 00:08:30,660 --> 00:08:31,910 They're minty gum. 198 00:08:31,910 --> 00:08:34,340 It doesn't make a difference. 199 00:08:34,340 --> 00:08:36,395 So basically the price is the same. 200 00:08:39,159 --> 00:08:41,320 If there's a supply shock, I don't know, they're made with 201 00:08:41,320 --> 00:08:41,890 the same shit. 202 00:08:41,890 --> 00:08:45,110 But let's say that Eclipse has some magic ingredient. 203 00:08:45,110 --> 00:08:47,010 And let's say the Eclipse magic ingredient got more 204 00:08:47,010 --> 00:08:50,540 expensive, so the supply curve shifted up. 205 00:08:50,540 --> 00:08:54,320 Well, Eclipse could not respond by raising its price. 206 00:08:54,320 --> 00:08:56,590 Because I just switched to Orbit. 207 00:08:56,590 --> 00:08:58,430 Or we often think of McDonald's and Burger King. 208 00:08:58,430 --> 00:08:59,290 Now, they're less perfect 209 00:08:59,290 --> 00:09:00,750 substitutes, but pretty perfect. 210 00:09:00,750 --> 00:09:02,730 If McDonald's started charging $10 for a hamburger, you 211 00:09:02,730 --> 00:09:04,130 wouldn't go there anymore. 212 00:09:04,130 --> 00:09:05,380 You'd go to Burger King. 213 00:09:08,560 --> 00:09:14,080 So if there's a supply shock to a provider that's facing a 214 00:09:14,080 --> 00:09:17,050 perfectly elastic demand curve, they cannot raise their 215 00:09:17,050 --> 00:09:19,160 price, because people will just switch. 216 00:09:19,160 --> 00:09:21,690 So quantity will fall a lot. 217 00:09:21,690 --> 00:09:25,970 Because if I'm supplying Eclipse gum, and it suddenly 218 00:09:25,970 --> 00:09:28,360 costs a lot more to produce Eclipse gum, but I can't raise 219 00:09:28,360 --> 00:09:32,340 my price, because I will lose all my business to Orbit, I'm 220 00:09:32,340 --> 00:09:34,510 just going to produce a lot less Eclipse. 221 00:09:34,510 --> 00:09:37,690 Because I'm losing money now. 222 00:09:37,690 --> 00:09:39,630 So with perfectly inelastic demand, the 223 00:09:39,630 --> 00:09:41,080 quantity didn't change. 224 00:09:41,080 --> 00:09:45,440 With perfectly elastic demand, we saw a big quantity change. 225 00:09:45,440 --> 00:09:50,090 So, more generally, what determines the quantity change 226 00:09:50,090 --> 00:09:53,430 in response to a price change is the elasticity. 227 00:09:53,430 --> 00:09:55,760 More generally, we're between these two cases of perfectly 228 00:09:55,760 --> 00:09:58,020 elastic and perfectly inelastic. 229 00:09:58,020 --> 00:10:02,720 And what's going to determine the price change is going to 230 00:10:02,720 --> 00:10:07,370 be the price elasticity of demand epsilon which is going 231 00:10:07,370 --> 00:10:13,560 to be the percentage change in quantity for each percentage 232 00:10:13,560 --> 00:10:19,998 change in price or, in calculus terms, dQ/dP. 233 00:10:22,830 --> 00:10:25,570 So it's, basically, the percentage change in quantity 234 00:10:25,570 --> 00:10:26,940 for the percentage change in price. 235 00:10:26,940 --> 00:10:33,620 So, for example, if for every 1% increase in price quantity 236 00:10:33,620 --> 00:10:37,990 falls 2%, that is a price elasticity of 237 00:10:37,990 --> 00:10:41,770 demand of minus 2. 238 00:10:41,770 --> 00:10:44,210 The price elasticity of demand is the percentage change in 239 00:10:44,210 --> 00:10:48,550 quantity for the percentage change in price. 240 00:10:48,550 --> 00:10:54,260 So inelastic demand is an epsilon of 0. 241 00:10:54,260 --> 00:10:58,200 There is no change in quantity when price changes. 242 00:10:58,200 --> 00:11:03,470 Perfectly elastic demand is an epsilon of negative infinity. 243 00:11:03,470 --> 00:11:06,400 Any epsilon change in price leads to a negative infinite 244 00:11:06,400 --> 00:11:07,190 change in quantity. 245 00:11:07,190 --> 00:11:08,560 Immediately, the quantity goes to 0 if you try 246 00:11:08,560 --> 00:11:10,250 to raise your price. 247 00:11:10,250 --> 00:11:12,072 So the price elasticity of demand will typically be 248 00:11:12,072 --> 00:11:14,270 between 0 and negative infinity. 249 00:11:14,270 --> 00:11:19,340 And the larger it is the more quantity will change when 250 00:11:19,340 --> 00:11:20,590 prices change. 251 00:11:25,240 --> 00:11:26,611 Questions about that? 252 00:11:26,611 --> 00:11:27,593 Yeah? 253 00:11:27,593 --> 00:11:31,030 AUDIENCE: So that formula, shouldn't it be dQ/dP times 254 00:11:31,030 --> 00:11:35,940 P/Q because dQ/dP just refers to the change of the quantity 255 00:11:35,940 --> 00:11:37,413 with respect to price, not 256 00:11:37,413 --> 00:11:39,400 necessarily the percent change. 257 00:11:42,066 --> 00:11:43,075 PROFESSOR: Yeah, you're right. 258 00:11:43,075 --> 00:11:44,520 I was trying to get too fancy with my calculus. 259 00:11:44,520 --> 00:11:45,010 You're right. 260 00:11:45,010 --> 00:11:48,500 Let's just stick with the non-calculus formula. 261 00:11:48,500 --> 00:11:50,740 I never should deviate from my notes. 262 00:11:50,740 --> 00:11:52,980 So let's just stick with the non-calculus formula. 263 00:11:52,980 --> 00:11:56,310 OK, other questions about this? 264 00:11:56,310 --> 00:11:56,950 OK. 265 00:11:56,950 --> 00:12:00,480 So, basically, that's the elasticity. 266 00:12:00,480 --> 00:12:03,700 That's going to be the elasticity. 267 00:12:03,700 --> 00:12:06,380 Now, an interesting point about elasticity is now, we're 268 00:12:06,380 --> 00:12:07,650 not going to get into producer theory 269 00:12:07,650 --> 00:12:08,990 for a couple of lectures. 270 00:12:08,990 --> 00:12:13,140 But as a little peek ahead about producer theory, let's 271 00:12:13,140 --> 00:12:15,920 think about how elasticity determines the money that 272 00:12:15,920 --> 00:12:17,930 producers make from selling their goods. 273 00:12:17,930 --> 00:12:23,820 Well, if a producer sells Q goods at a price P, they make 274 00:12:23,820 --> 00:12:30,280 revenues R. Revenues are the price times the quantity. 275 00:12:30,280 --> 00:12:37,140 The amount of money a producer makes when it sells goods, its 276 00:12:37,140 --> 00:12:39,300 revenues, this isn't its profits. 277 00:12:39,300 --> 00:12:39,880 We're not having profits. 278 00:12:39,880 --> 00:12:41,290 It's just the amount of money it makes, not the amount of 279 00:12:41,290 --> 00:12:42,430 money it takes home at the end of the day. 280 00:12:42,430 --> 00:12:44,400 I'm ignoring the cost of making the goods. 281 00:12:44,400 --> 00:12:45,840 The amount of total revenues it makes 282 00:12:45,840 --> 00:12:48,210 is price times quantity. 283 00:12:48,210 --> 00:12:51,860 Well, we can then say that the change in revenues with 284 00:12:51,860 --> 00:12:55,610 respect to price is what? 285 00:12:55,610 --> 00:13:02,600 It's Q plus dQ, plus delta Q-- let me put it this way to make 286 00:13:02,600 --> 00:13:03,720 my math clearer-- 287 00:13:03,720 --> 00:13:10,260 plus P times delta Q over delta P. That's how revenues 288 00:13:10,260 --> 00:13:12,840 change with respect to price. 289 00:13:12,840 --> 00:13:14,810 Or, in other words, plugging in from the elasticity 290 00:13:14,810 --> 00:13:25,965 formula, delta R over delta P equals Q times 1 plus epsilon. 291 00:13:29,770 --> 00:13:34,620 So, in other words, what this says is that if you're a 292 00:13:34,620 --> 00:13:37,500 producer, and you're trying to decide whether to raise your 293 00:13:37,500 --> 00:13:40,360 price, whether that will increase revenues, it all 294 00:13:40,360 --> 00:13:42,420 depends on the elasticity. 295 00:13:42,420 --> 00:13:48,150 If the elasticity is between 0 and minus 1, then raising 296 00:13:48,150 --> 00:13:51,120 prices will raise revenues. 297 00:13:51,120 --> 00:13:55,310 If the elasticity is greater than minus 1, then raising 298 00:13:55,310 --> 00:13:58,900 prices will lower revenues. 299 00:13:58,900 --> 00:14:02,230 We're often faced with the issue of why did they charge 300 00:14:02,230 --> 00:14:04,040 this much for this good, or should they raise their prices 301 00:14:04,040 --> 00:14:05,270 or not raise their price. 302 00:14:05,270 --> 00:14:08,540 Well, that's all about the elasticity of demand. 303 00:14:08,540 --> 00:14:11,710 The elasticity of demand will determine whether they're 304 00:14:11,710 --> 00:14:14,290 going to make more money by raising their price or lose 305 00:14:14,290 --> 00:14:15,820 money by raising their price. 306 00:14:15,820 --> 00:14:18,410 For Eclipse gum, their elasticity of demand is well 307 00:14:18,410 --> 00:14:20,650 above minus 1 in absolute value, so they're going to 308 00:14:20,650 --> 00:14:21,940 lose money by raising their price. 309 00:14:21,940 --> 00:14:24,610 If they take the current level of Eclipse, for every penny 310 00:14:24,610 --> 00:14:27,260 they raise, they'll lose money. 311 00:14:27,260 --> 00:14:28,790 For insulin, for every penny they raise, 312 00:14:28,790 --> 00:14:30,790 they'll make money. 313 00:14:30,790 --> 00:14:33,100 And then you might say well, then how come the price of 314 00:14:33,100 --> 00:14:34,950 insulin isn't infinity and the price of 315 00:14:34,950 --> 00:14:36,240 Eclipse gum isn't zero? 316 00:14:36,240 --> 00:14:37,780 Well, that's what we'll talk about in a few weeks. 317 00:14:37,780 --> 00:14:41,360 Because it also depends on the costs of producing it. 318 00:14:41,360 --> 00:14:43,890 But at the end of the day, that's what's going to 319 00:14:43,890 --> 00:14:45,990 determine the money that's made by producers when they 320 00:14:45,990 --> 00:14:48,860 change their prices. 321 00:14:48,860 --> 00:14:50,110 Questions about that? 322 00:14:53,120 --> 00:14:53,790 OK. 323 00:14:53,790 --> 00:14:57,880 So now, that's how we think about the shape 324 00:14:57,880 --> 00:14:59,870 of supply and demand. 325 00:14:59,870 --> 00:15:01,600 The shape of supply and demand is determined by these 326 00:15:01,600 --> 00:15:03,360 elasticities. 327 00:15:03,360 --> 00:15:06,920 So now we have to get into OK, well, where do we get these 328 00:15:06,920 --> 00:15:08,860 elasticities from? 329 00:15:08,860 --> 00:15:20,870 And that is the main topic of empirical economics which is 330 00:15:20,870 --> 00:15:25,370 estimating these kinds of elasticities, estimating these 331 00:15:25,370 --> 00:15:26,620 types of elasticities. 332 00:15:32,120 --> 00:15:36,220 So one of the first distinctions I drew in the 333 00:15:36,220 --> 00:15:37,900 lectures is between theoretical economics and 334 00:15:37,900 --> 00:15:39,520 empirical economics. 335 00:15:39,520 --> 00:15:44,160 Theoretical economics can tell us this is what a graph looks 336 00:15:44,160 --> 00:15:46,620 like and supply and demand. 337 00:15:49,280 --> 00:15:51,140 Theoretical economics can't really tell us how big, for 338 00:15:51,140 --> 00:15:53,360 example, an elasticity is going to be. 339 00:15:53,360 --> 00:15:55,400 It can tell us, there's more substitutes or less 340 00:15:55,400 --> 00:15:57,840 substitutes so we can rank them. 341 00:15:57,840 --> 00:16:00,450 We know the elasticity for Eclipse gum has got to be 342 00:16:00,450 --> 00:16:03,620 higher than the elasticity for insulin. 343 00:16:03,620 --> 00:16:06,620 But from the theoretical model, we can't say what the 344 00:16:06,620 --> 00:16:08,660 elasticity actually is. 345 00:16:08,660 --> 00:16:11,470 To say what an elasticity actually is, we need to go to 346 00:16:11,470 --> 00:16:12,870 an empirical model. 347 00:16:12,870 --> 00:16:15,940 We actually need to bring data to bear on the question. 348 00:16:15,940 --> 00:16:17,840 And this is very difficult. 349 00:16:17,840 --> 00:16:20,680 Because here we face the fundamental conundrum facing 350 00:16:20,680 --> 00:16:26,220 the empirical economist which is distinguishing causation 351 00:16:26,220 --> 00:16:27,470 from correlation. 352 00:16:30,400 --> 00:16:33,840 And the whole guts of empirical economics is all 353 00:16:33,840 --> 00:16:37,490 about this question, distinguishing causation from 354 00:16:37,490 --> 00:16:38,740 correlation. 355 00:16:40,800 --> 00:16:43,895 The classic story that illustrates this, it's due to 356 00:16:43,895 --> 00:16:46,130 my colleague, Frank Fisher, from a textbook many years 357 00:16:46,130 --> 00:16:50,420 ago, was the story of in ancient Russia there was a 358 00:16:50,420 --> 00:16:54,390 cholera outbreak, and many people were dying. 359 00:16:54,390 --> 00:16:57,840 So the government decided to send doctors out to try to 360 00:16:57,840 --> 00:16:59,470 solve the problem. 361 00:16:59,470 --> 00:17:01,850 And where there were more people sick, 362 00:17:01,850 --> 00:17:04,030 they sent more doctors. 363 00:17:04,030 --> 00:17:06,190 Well, the peasants said, wait a second. 364 00:17:06,190 --> 00:17:09,420 We observe that where there's more doctors, more people are 365 00:17:09,420 --> 00:17:10,510 dying from cholera. 366 00:17:10,510 --> 00:17:12,700 So the doctors must be causing the cholera. 367 00:17:12,700 --> 00:17:16,190 So they rose up and killed the doctors. 368 00:17:16,190 --> 00:17:21,250 The peasants confused causation with correlation. 369 00:17:21,250 --> 00:17:23,190 They thought that the fact that you saw more people dying 370 00:17:23,190 --> 00:17:24,800 where there's more doctors meant that doctors were 371 00:17:24,800 --> 00:17:25,455 causing the disease. 372 00:17:25,455 --> 00:17:26,310 Clearly that's wrong. 373 00:17:26,310 --> 00:17:27,560 That's why they were peasants. 374 00:17:29,550 --> 00:17:31,435 But it's not just peasants that make this mistake. 375 00:17:35,170 --> 00:17:39,070 For example, in 1988, Harvard University, our illustrious 376 00:17:39,070 --> 00:17:41,525 neighbor to the south, I guess, west, 377 00:17:41,525 --> 00:17:42,180 east, I don't know. 378 00:17:42,180 --> 00:17:42,750 Which way is Harvard? 379 00:17:42,750 --> 00:17:44,300 I don't know directions, down the street. 380 00:17:46,930 --> 00:17:49,020 A Harvard University dean conducted an interview with a 381 00:17:49,020 --> 00:17:50,820 set of freshmen. 382 00:17:50,820 --> 00:17:53,140 And they found that those that had taken 383 00:17:53,140 --> 00:17:56,030 SAT preparation courses-- 384 00:17:56,030 --> 00:17:57,890 now, you all took SAT preparation courses. 385 00:17:57,890 --> 00:18:00,650 But in 1988, not everyone did. 386 00:18:00,650 --> 00:18:04,310 Those who'd taken SAT preparation courses scored an 387 00:18:04,310 --> 00:18:07,460 average of 63 points lower-- 388 00:18:07,460 --> 00:18:10,070 this was back when the SAT was 1600 points-- 389 00:18:10,070 --> 00:18:13,100 63 points lower on their SATs then those that had not taken 390 00:18:13,100 --> 00:18:15,620 preparation courses. 391 00:18:15,620 --> 00:18:18,690 The dean concluded that preparation courses were 392 00:18:18,690 --> 00:18:22,600 unhelpful, and that the testing industry was preying 393 00:18:22,600 --> 00:18:24,590 on the insecurities of students to 394 00:18:24,590 --> 00:18:27,800 provide a useless service. 395 00:18:27,800 --> 00:18:31,190 Why was the dean confusing causation with correlation? 396 00:18:31,190 --> 00:18:33,648 What did the dean get wrong in drawing that conclusion? 397 00:18:33,648 --> 00:18:34,136 Yeah? 398 00:18:34,136 --> 00:18:38,040 AUDIENCE: What had probably happened is the students who 399 00:18:38,040 --> 00:18:40,968 got worse scores realized that they wanted to try and improve 400 00:18:40,968 --> 00:18:42,920 their scores by taking an SAT prep class. 401 00:18:42,920 --> 00:18:46,336 So that's why there is a lower average score for the people 402 00:18:46,336 --> 00:18:47,312 who had taken the class. 403 00:18:47,312 --> 00:18:49,790 PROFESSOR: Generally, the people who needed the help the 404 00:18:49,790 --> 00:18:51,660 most took the most courses. 405 00:18:51,660 --> 00:18:55,330 And so they had an underlying lower score. 406 00:18:55,330 --> 00:18:58,620 So, in fact, you can't tell anything from the fact that 407 00:18:58,620 --> 00:19:00,805 the people who took the prep course scored worse. 408 00:19:05,090 --> 00:19:08,300 It's just another excellent example of confusing causation 409 00:19:08,300 --> 00:19:11,350 with correlation. 410 00:19:11,350 --> 00:19:13,330 And that's another example. 411 00:19:13,330 --> 00:19:16,640 Another example I like quite a lot is studies of 412 00:19:16,640 --> 00:19:19,200 breastfeeding. 413 00:19:19,200 --> 00:19:22,760 There are numbers of studies of breastfeeding, especially 414 00:19:22,760 --> 00:19:27,640 in developing countries, where they found that the longer 415 00:19:27,640 --> 00:19:31,720 children were breastfed the sicker they were. 416 00:19:31,720 --> 00:19:32,540 So they concluded that 417 00:19:32,540 --> 00:19:35,540 breastfeeding was bad for kids. 418 00:19:35,540 --> 00:19:37,140 Well, that's not the truth. 419 00:19:37,140 --> 00:19:40,650 The truth is the sicker kids need to be breastfed more, 420 00:19:40,650 --> 00:19:43,860 because breastfeeding is actually good for kids. 421 00:19:43,860 --> 00:19:46,320 And they just confused the causation with the 422 00:19:46,320 --> 00:19:47,550 correlation. 423 00:19:47,550 --> 00:19:49,040 Now, these are all fun examples. 424 00:19:49,040 --> 00:19:53,970 But the truth is this is a common mistake made by 425 00:19:53,970 --> 00:19:56,880 citizens, policy makers, everyone in the real world. 426 00:19:56,880 --> 00:19:59,210 It's taking two things that move together and assuming one 427 00:19:59,210 --> 00:20:01,110 causes the other. 428 00:20:01,110 --> 00:20:04,400 And this is the fundamental conundrum facing empirical 429 00:20:04,400 --> 00:20:08,010 economics in trying to address these kinds of things like 430 00:20:08,010 --> 00:20:11,170 measuring elasticities. 431 00:20:11,170 --> 00:20:15,120 So to understand that, let's think about the issue of 432 00:20:15,120 --> 00:20:19,150 trying to estimate the elasticity of demand for pork. 433 00:20:19,150 --> 00:20:21,950 Let's say you have the exciting job of estimating the 434 00:20:21,950 --> 00:20:24,130 elasticity of demand for pork. 435 00:20:24,130 --> 00:20:26,020 That's your assignment. 436 00:20:26,020 --> 00:20:29,610 Well, you say, wait a second. 437 00:20:29,610 --> 00:20:35,920 What we learned in class, as shown in Figure 3-4, is that 438 00:20:35,920 --> 00:20:42,420 the price of pork can rise for very different reasons. 439 00:20:42,420 --> 00:20:49,810 Figure 3-4, we start at an initial equilibrium like E1 440 00:20:49,810 --> 00:20:51,650 with a quantity like Q1 and a price P1. 441 00:20:54,510 --> 00:21:00,160 Now, imagine that there was a shift in demand, because the 442 00:21:00,160 --> 00:21:01,390 price of beef rose, remember? 443 00:21:01,390 --> 00:21:02,510 The price of beef rose. 444 00:21:02,510 --> 00:21:04,750 That shifted demand from D1 to D2. 445 00:21:04,750 --> 00:21:06,690 What did that lead to? 446 00:21:06,690 --> 00:21:11,050 A higher price and a higher quantity. 447 00:21:11,050 --> 00:21:13,660 So if you took that diagram-- 448 00:21:13,660 --> 00:21:15,730 forget the supply shift for a minute, just imagine that's 449 00:21:15,730 --> 00:21:17,100 the change-- 450 00:21:17,100 --> 00:21:18,150 and you said, aha. 451 00:21:18,150 --> 00:21:20,980 I can measure the elasticity. 452 00:21:20,980 --> 00:21:24,160 I see here there's a change in price. 453 00:21:24,160 --> 00:21:26,280 I can then look at how quantity changed. 454 00:21:26,280 --> 00:21:27,650 And I'll get the elasticity right after all. 455 00:21:27,650 --> 00:21:30,250 It's delta Q over Q or delta P over P. 456 00:21:30,250 --> 00:21:35,440 So I just look, and I take Q2 prime minus Q1 over Q1. 457 00:21:35,440 --> 00:21:38,490 That's the percentage change in Q. I take P2 458 00:21:38,490 --> 00:21:39,710 over P1 over P1. 459 00:21:39,710 --> 00:21:40,780 That's the percentage change in price. 460 00:21:40,780 --> 00:21:42,750 And what do I get? 461 00:21:42,750 --> 00:21:44,980 A wrong signed elasticity is what I get. 462 00:21:44,980 --> 00:21:47,880 I get a positive elasticity, because Q is going up 463 00:21:47,880 --> 00:21:50,930 and P is going up. 464 00:21:50,930 --> 00:21:51,780 Why? 465 00:21:51,780 --> 00:21:56,140 Because I'm confusing causation with correlation. 466 00:21:56,140 --> 00:22:01,920 It's not the price change that caused quantity to change. 467 00:22:01,920 --> 00:22:03,240 In fact, it's the opposite. 468 00:22:03,240 --> 00:22:07,560 It's a taste shift, which caused quantity to increase 469 00:22:07,560 --> 00:22:10,450 which drove up the price. 470 00:22:10,450 --> 00:22:14,460 It was a demand increase which caused the quantity demanded 471 00:22:14,460 --> 00:22:16,320 to increase which drove up the price. 472 00:22:16,320 --> 00:22:18,310 So it's the quantity driving the price, not the price 473 00:22:18,310 --> 00:22:19,460 driving the quantity. 474 00:22:19,460 --> 00:22:21,630 So if you looked at that simple example, as many people 475 00:22:21,630 --> 00:22:23,680 in the real world do, they'd say, hey, look. 476 00:22:23,680 --> 00:22:26,550 Higher prices cause higher quantities. 477 00:22:26,550 --> 00:22:28,220 You're getting the wrong answer. 478 00:22:28,220 --> 00:22:31,190 Because you're confusing correlation which is the 479 00:22:31,190 --> 00:22:33,020 higher price is correlated with the higher quantity. 480 00:22:33,020 --> 00:22:35,420 Because there was a common factor causing both of them 481 00:22:35,420 --> 00:22:39,620 which is the demand shift and not causation. 482 00:22:39,620 --> 00:22:41,965 The higher price did not cause the higher quantity. 483 00:22:44,570 --> 00:22:46,270 What do we need to do? 484 00:22:46,270 --> 00:22:50,440 We need to distinguish why the price increased. 485 00:22:50,440 --> 00:22:51,940 We need to distinguish why the price 486 00:22:51,940 --> 00:22:55,950 increased to measure this. 487 00:22:55,950 --> 00:23:03,360 If, instead, we looked at a shift in supply such as the 488 00:23:03,360 --> 00:23:06,680 case that's shifting from S1 to S2 and moving the 489 00:23:06,680 --> 00:23:10,150 equilibrium from E1 to E2, then you would 490 00:23:10,150 --> 00:23:11,960 get the right answer. 491 00:23:11,960 --> 00:23:13,210 Because then you'd say, look. 492 00:23:13,210 --> 00:23:15,610 Something independent to consumers 493 00:23:15,610 --> 00:23:17,290 shifted up the price. 494 00:23:17,290 --> 00:23:19,970 Some shock to the supply of pork shifted up the price. 495 00:23:19,970 --> 00:23:22,710 And we saw that their quantity fell as a result. 496 00:23:25,530 --> 00:23:26,610 What's the key? 497 00:23:26,610 --> 00:23:30,650 The key is that to measure an elasticity of demand, you're 498 00:23:30,650 --> 00:23:33,170 measuring the slope of the demand curve. 499 00:23:33,170 --> 00:23:36,830 So you need to shift along a demand curve, not shift the 500 00:23:36,830 --> 00:23:38,370 demand curve itself. 501 00:23:38,370 --> 00:23:41,990 So if you look at this figure, what's the concept we want? 502 00:23:41,990 --> 00:23:44,830 We want the slope of the demand curve. 503 00:23:44,830 --> 00:23:49,750 Well, you get that by shifting from E1 to E2, because you 504 00:23:49,750 --> 00:23:51,730 shift along the demand curve. 505 00:23:51,730 --> 00:23:55,040 So by looking at what happens to quantity as price rises 506 00:23:55,040 --> 00:23:57,540 from E1 to E2, you get the slope of the demand curve. 507 00:23:57,540 --> 00:24:00,105 You get that delta Q over delta P you want. 508 00:24:00,105 --> 00:24:03,010 But from E1 to E2 prime, you're not shifting along the 509 00:24:03,010 --> 00:24:04,360 demand curve. 510 00:24:04,360 --> 00:24:08,410 You're actually measuring the elasticity of supply. 511 00:24:08,410 --> 00:24:09,272 You're measuring the elasticity of supply. 512 00:24:09,272 --> 00:24:12,040 You're shifting along a supply curve. 513 00:24:12,040 --> 00:24:15,640 So you're actually answering a different question, a relevant 514 00:24:15,640 --> 00:24:17,020 question, but a different one. 515 00:24:17,020 --> 00:24:20,750 That question is, what's the elasticity of supply? 516 00:24:20,750 --> 00:24:25,080 How willing are pork producers to supply pork as 517 00:24:25,080 --> 00:24:26,100 the price goes up? 518 00:24:26,100 --> 00:24:30,470 So it's the same delta Q over delta P. But here we did the 519 00:24:30,470 --> 00:24:32,050 elasticity of demand. 520 00:24:32,050 --> 00:24:36,500 There's a corresponding elasticity of supply which is 521 00:24:36,500 --> 00:24:39,590 measured the same way. 522 00:24:39,590 --> 00:24:41,865 It's delta Q over delta P, but it's for a 523 00:24:41,865 --> 00:24:43,460 different kind of shock. 524 00:24:43,460 --> 00:24:45,780 It's what you get from moving along the supply curve. 525 00:24:45,780 --> 00:24:49,940 So if we went from E1 to E2 prime, we can use that to 526 00:24:49,940 --> 00:24:52,540 measure the elasticity of supply or the slope of the 527 00:24:52,540 --> 00:24:54,890 supply curve. 528 00:24:54,890 --> 00:24:58,230 And we do that if something shifts demand to move us along 529 00:24:58,230 --> 00:24:59,420 the supply curve. 530 00:24:59,420 --> 00:25:04,250 From E1 to E2, we measure the elasticity of demand as 531 00:25:04,250 --> 00:25:06,130 something shifts supply and moves us 532 00:25:06,130 --> 00:25:10,390 along the demand curve. 533 00:25:10,390 --> 00:25:14,140 So what we need to measure the elasticity of demand is 534 00:25:14,140 --> 00:25:16,880 something which shifts supply but does not, 535 00:25:16,880 --> 00:25:19,930 itself, affect demand. 536 00:25:19,930 --> 00:25:23,790 And the best example of this that we use in economics, a 537 00:25:23,790 --> 00:25:28,610 great example, is government policy which comes along and 538 00:25:28,610 --> 00:25:31,630 changes the supply conditions for a good. 539 00:25:31,630 --> 00:25:35,420 So, for example, let's think about a tax on pork. 540 00:25:35,420 --> 00:25:40,290 So if you go to Figure 3-5, imagine the government came 541 00:25:40,290 --> 00:25:43,970 along and taxed pork. 542 00:25:46,900 --> 00:25:49,040 The government comes along and taxes pork. 543 00:25:49,040 --> 00:25:53,240 Let's think about what a tax on pork does. 544 00:25:53,240 --> 00:25:55,570 The government comes along, and let's say the pork market 545 00:25:55,570 --> 00:26:01,300 is initially in equilibrium at $3.30 with 220 million 546 00:26:01,300 --> 00:26:04,360 kilograms of pork sold. 547 00:26:04,360 --> 00:26:08,610 Now the government comes along and says that it's going to 548 00:26:08,610 --> 00:26:14,500 charge $1.05 in tax for every kilogram of pork. 549 00:26:14,500 --> 00:26:19,360 So it's going to impose a tax of $1.05 550 00:26:19,360 --> 00:26:21,990 per kilogram on producers. 551 00:26:21,990 --> 00:26:24,350 So it's saying to producers of pork, for every kilogram of 552 00:26:24,350 --> 00:26:27,270 pork you sell, you have to send a check to the 553 00:26:27,270 --> 00:26:31,120 government for $1.05. 554 00:26:31,120 --> 00:26:32,900 For every kilogram of pork you sell, you have to send a check 555 00:26:32,900 --> 00:26:36,750 to the government for $1.05. 556 00:26:36,750 --> 00:26:41,520 Now, somebody talk me through how a supplier 557 00:26:41,520 --> 00:26:43,060 thinks through that. 558 00:26:43,060 --> 00:26:44,390 How does a supplier react to that? 559 00:26:44,390 --> 00:26:45,518 What do they think? 560 00:26:45,518 --> 00:26:51,030 They're initially happy at E1 selling 220 million 561 00:26:51,030 --> 00:26:53,450 kilograms at $3.30. 562 00:26:53,450 --> 00:26:56,150 What happens when the government comes in and says 563 00:26:56,150 --> 00:26:58,570 you have to pay $1.05 for every 564 00:26:58,570 --> 00:27:00,230 kilogram of pork you sell? 565 00:27:00,230 --> 00:27:02,930 What happens? 566 00:27:02,930 --> 00:27:03,752 Yeah. 567 00:27:03,752 --> 00:27:06,282 AUDIENCE: The producer decides that the current amount of 568 00:27:06,282 --> 00:27:10,018 money they have will not be able to buy as much inputs to 569 00:27:10,018 --> 00:27:11,464 create their products. 570 00:27:11,464 --> 00:27:12,265 So they can produce less. 571 00:27:12,265 --> 00:27:12,570 PROFESSOR: Exactly. 572 00:27:12,570 --> 00:27:16,020 So, in other words, the cost of producing just rose. 573 00:27:16,020 --> 00:27:19,110 So what do they do? 574 00:27:19,110 --> 00:27:23,060 So, in other words, what they say is look, effectively, if I 575 00:27:23,060 --> 00:27:30,900 was happy before selling 220 million kilograms at $3.30, to 576 00:27:30,900 --> 00:27:35,330 keep me equally happy selling 220 million kilograms, I'm 577 00:27:35,330 --> 00:27:36,340 going to have to raise the price. 578 00:27:36,340 --> 00:27:38,300 We should add this to graph, actually. 579 00:27:38,300 --> 00:27:42,060 If you draw a vertical line up for me, one to the S2 curve. 580 00:27:42,060 --> 00:27:44,420 Draw a little dashed line up from the E1 to the S2 curve 581 00:27:44,420 --> 00:27:45,730 and then over. 582 00:27:45,730 --> 00:27:51,030 That price intersection will be $4.35. 583 00:27:51,030 --> 00:27:54,330 So in other words, if you want me to keep producing 220 584 00:27:54,330 --> 00:27:57,490 million kilograms of pork, I'm going to have 585 00:27:57,490 --> 00:28:00,856 to get $4.35 a kilogram. 586 00:28:00,856 --> 00:28:03,120 And you might say, what gives you the right to get that? 587 00:28:03,120 --> 00:28:04,110 And it's not about rights. 588 00:28:04,110 --> 00:28:07,320 It's about what producers are willing to do. 589 00:28:07,320 --> 00:28:11,770 That same mathematics, that same supply curve that tells 590 00:28:11,770 --> 00:28:16,050 us they're willing to sell 220 million kilograms at $3.30 591 00:28:16,050 --> 00:28:18,460 says, if you want them to keep selling 220 million kilograms 592 00:28:18,460 --> 00:28:21,190 but also pay $1.05 to the government, they're going to 593 00:28:21,190 --> 00:28:24,820 have to get $4.35 a kilogram. 594 00:28:24,820 --> 00:28:28,160 So what happens is that's a supply shift. 595 00:28:28,160 --> 00:28:30,060 And with the same reaction we saw last time with the 596 00:28:30,060 --> 00:28:34,180 drought, the price goes up, consumers demand less, and you 597 00:28:34,180 --> 00:28:37,910 reach a new equilibrium at the price E2. 598 00:28:37,910 --> 00:28:43,060 You reach a new equilibrium where you sell 206 million 599 00:28:43,060 --> 00:28:49,140 kilograms for a price of $4.00. 600 00:28:49,140 --> 00:28:51,840 So someone tell me how I use this example to find 601 00:28:51,840 --> 00:28:53,303 elasticity of demand. 602 00:28:53,303 --> 00:28:55,275 Yeah. 603 00:28:55,275 --> 00:28:58,233 AUDIENCE: I guess you need to know that the change in price 604 00:28:58,233 --> 00:29:00,205 traveled along the demand curve. 605 00:29:00,205 --> 00:29:02,177 So you know that it's not [INAUDIBLE PHRASE]. 606 00:29:05,135 --> 00:29:06,614 PROFESSOR: OK, so tell me. 607 00:29:06,614 --> 00:29:08,093 You don't have to do the math in your head. 608 00:29:08,093 --> 00:29:09,572 But how would I compute it? 609 00:29:09,572 --> 00:29:12,462 AUDIENCE: You would take E1 and E2, and then you would do 610 00:29:12,462 --> 00:29:14,032 the price over the quantity change. 611 00:29:14,032 --> 00:29:15,550 PROFESSOR: Right, exactly. 612 00:29:15,550 --> 00:29:21,835 So the quantity change delta Q over Q, is what? 613 00:29:25,110 --> 00:29:26,960 It's minus 14 over 220. 614 00:29:26,960 --> 00:29:30,680 It fell by 14 million kilograms over 220. 615 00:29:30,680 --> 00:29:39,990 The price change, delta P over P, the price rose 616 00:29:39,990 --> 00:29:41,320 from $3.30 to $4.00. 617 00:29:41,320 --> 00:29:49,370 So the price change is $0.70 over $3.30. 618 00:29:49,370 --> 00:29:55,330 And using those, you end up with a price 619 00:29:55,330 --> 00:29:59,900 elasticity of minus 0.3. 620 00:29:59,900 --> 00:30:06,790 Or, in other words, there's a 6.4% change in quantity. 621 00:30:06,790 --> 00:30:15,490 This is minus 6.4% for a 21% change in price. 622 00:30:15,490 --> 00:30:20,350 So quantity falls by 6.4% when price goes up by 21%. 623 00:30:20,350 --> 00:30:23,580 That's a price elasticity of minus 0.3. 624 00:30:23,580 --> 00:30:27,380 Or that's a relatively inelastic demand. 625 00:30:27,380 --> 00:30:29,950 It's not perfectly inelastic, but it's relatively inelastic. 626 00:30:29,950 --> 00:30:32,900 In other words, at that point, pork producers could make 627 00:30:32,900 --> 00:30:35,455 money by raising the price. 628 00:30:35,455 --> 00:30:37,760 Now, you might say well, why didn't they? 629 00:30:37,760 --> 00:30:41,560 That's something we'll discuss in a couple weeks. 630 00:30:41,560 --> 00:30:43,870 But at that point, demand is relatively inelastic. 631 00:30:43,870 --> 00:30:46,020 And you've got a convincing estimate, because you moved 632 00:30:46,020 --> 00:30:47,190 along that demand curve. 633 00:30:47,190 --> 00:30:49,750 You used the supply shift. 634 00:30:49,750 --> 00:30:52,920 Now, we're going to talk about taxation much, much later in 635 00:30:52,920 --> 00:30:54,720 the semester. 636 00:30:54,720 --> 00:30:56,780 Let me just talk for one minute about what we learned 637 00:30:56,780 --> 00:30:57,550 from this graph. 638 00:30:57,550 --> 00:30:58,670 What happens? 639 00:30:58,670 --> 00:31:02,300 Well, the shaded area is the money the government raises 640 00:31:02,300 --> 00:31:04,240 from its tax. 641 00:31:04,240 --> 00:31:09,490 The government has a tax of $1.05 at 206 million 642 00:31:09,490 --> 00:31:14,810 kilograms. So it raises $1.05 times 206 million kilograms 643 00:31:14,810 --> 00:31:16,660 which is that shaded area. 644 00:31:16,660 --> 00:31:18,400 There are two points to note the we'll come back to later 645 00:31:18,400 --> 00:31:19,415 in the semester. 646 00:31:19,415 --> 00:31:22,740 The first point to note is the amount of money the government 647 00:31:22,740 --> 00:31:27,840 raises will depend directly on the elasticity of demand. 648 00:31:27,840 --> 00:31:32,420 Can anyone tell me how much money the government would 649 00:31:32,420 --> 00:31:37,540 raise if you had a perfectly inelastic demand? 650 00:31:40,590 --> 00:31:41,043 Yeah. 651 00:31:41,043 --> 00:31:42,293 AUDIENCE: [INAUDIBLE PHRASE]. 652 00:31:45,150 --> 00:31:47,640 PROFESSOR: Right. 653 00:31:47,640 --> 00:31:51,520 If we think about this demand curve being perfectly flat, if 654 00:31:51,520 --> 00:31:54,070 we think about this demand curve being perfectly flat, 655 00:31:54,070 --> 00:32:01,700 then basically the producer can't charge any more for 656 00:32:01,700 --> 00:32:03,310 their good. 657 00:32:03,310 --> 00:32:07,290 So it's going to depend on whether the producer is 658 00:32:07,290 --> 00:32:11,510 willing to sell at $1.05 less and how much less they're 659 00:32:11,510 --> 00:32:11,870 willing to sell. 660 00:32:11,870 --> 00:32:12,992 If they're willing to sell a lot less, they're going to 661 00:32:12,992 --> 00:32:15,790 make a lot less money. 662 00:32:15,790 --> 00:32:19,430 It's going to be where that second supply curve intersects 663 00:32:19,430 --> 00:32:20,820 a flat demand curve. 664 00:32:20,820 --> 00:32:22,240 So that quantity is going to be a lot smaller. 665 00:32:22,240 --> 00:32:23,580 We don't have it on the diagram. 666 00:32:23,580 --> 00:32:29,230 But you see where that dashed line at $3.30 intersects S2, 667 00:32:29,230 --> 00:32:30,500 that's way to the left. 668 00:32:30,500 --> 00:32:32,690 Quantity is going to fall a ton in this market. 669 00:32:32,690 --> 00:32:34,690 When quantity falls, the government is going to raise a 670 00:32:34,690 --> 00:32:36,170 lot less money. 671 00:32:36,170 --> 00:32:37,750 Because the government raises $1.05 on every 672 00:32:37,750 --> 00:32:39,660 unit sold at the end. 673 00:32:39,660 --> 00:32:44,180 So if the government taxes very elastically demanded 674 00:32:44,180 --> 00:32:46,410 goods, it's going to raise less money. 675 00:32:46,410 --> 00:32:49,710 If it taxes inelastically demanded goods like insulin, 676 00:32:49,710 --> 00:32:51,540 it's going to raise more money, because the quantity 677 00:32:51,540 --> 00:32:53,550 doesn't change. 678 00:32:53,550 --> 00:32:53,810 Yeah. 679 00:32:53,810 --> 00:32:56,440 AUDIENCE: So cigarettes are relatively inelastic. 680 00:32:56,440 --> 00:32:57,570 PROFESSOR: Yes, exactly. 681 00:32:57,570 --> 00:32:58,830 Cigarettes are relatively inelastic. 682 00:32:58,830 --> 00:33:01,790 The elasticity is around minus 0.5. 683 00:33:01,790 --> 00:33:05,350 So the government will actually raise money by 684 00:33:05,350 --> 00:33:06,670 raising the cigarette tax. 685 00:33:12,180 --> 00:33:14,670 Those of us, as good liberals, think we should tax yachts. 686 00:33:14,670 --> 00:33:15,475 Let's tax yachts. 687 00:33:15,475 --> 00:33:17,640 Only rich guy have yachts. 688 00:33:17,640 --> 00:33:18,580 The problem is yachts are 689 00:33:18,580 --> 00:33:20,730 incredibly elastically demanded. 690 00:33:20,730 --> 00:33:24,130 So you raise a lot less money taxing yachts than you think. 691 00:33:24,130 --> 00:33:28,020 Because guys buy fewer yachts, and you don't raise as much 692 00:33:28,020 --> 00:33:28,640 money as you think you would. 693 00:33:28,640 --> 00:33:31,270 You still raise some, and it still may be worth it. 694 00:33:31,270 --> 00:33:33,510 But you raise less than you think. 695 00:33:33,510 --> 00:33:39,200 So that's one sort of observation about this. 696 00:33:39,200 --> 00:33:41,800 It's basically how much money you'll raise will be a 697 00:33:41,800 --> 00:33:47,320 function of how elastic the demand is. 698 00:33:47,320 --> 00:33:50,910 The other important observation to make is why 699 00:33:50,910 --> 00:33:53,280 it's actually hard for governments to figure out how 700 00:33:53,280 --> 00:33:55,700 much money they're going to raise for a tax. 701 00:33:55,700 --> 00:33:58,730 Because, to figure it out, they need to know these 702 00:33:58,730 --> 00:33:59,400 elasticities. 703 00:33:59,400 --> 00:34:01,860 That is, the naive thing to do would have been to say what? 704 00:34:01,860 --> 00:34:04,790 Well, we're selling 220 million kilograms of pork. 705 00:34:04,790 --> 00:34:06,080 That's $1.05. 706 00:34:06,080 --> 00:34:07,240 We're going to tax each kilogram. 707 00:34:07,240 --> 00:34:08,560 So that's 220 million times $1.05. 708 00:34:08,560 --> 00:34:10,830 And that's how much money we raise. 709 00:34:10,830 --> 00:34:12,540 Well, that's wrong, we know, because that 710 00:34:12,540 --> 00:34:14,659 assumes inelastic demand. 711 00:34:14,659 --> 00:34:16,699 If demand's elastic, they'll raise less than that. 712 00:34:16,699 --> 00:34:18,389 Well, if we want to figure out how much a government is going 713 00:34:18,389 --> 00:34:19,800 to raise from a tax, they've got to know what these 714 00:34:19,800 --> 00:34:20,639 elasticities are. 715 00:34:20,639 --> 00:34:22,754 And those are actually pretty hard things to know. 716 00:34:22,754 --> 00:34:24,620 So that's why there's uncertainty. 717 00:34:24,620 --> 00:34:27,860 That's why when politicians will say, this tax will raise 718 00:34:27,860 --> 00:34:31,360 x and you'll hear the New York Times report, the tax will 719 00:34:31,360 --> 00:34:35,030 raise x, that is a guess. 720 00:34:35,030 --> 00:34:36,850 Those are guesses, because they depend on our best 721 00:34:36,850 --> 00:34:39,530 estimate of the key elasticities that determine 722 00:34:39,530 --> 00:34:40,600 how people respond. 723 00:34:40,600 --> 00:34:40,850 Yeah. 724 00:34:40,850 --> 00:34:41,962 AUDIENCE: But in Washington you have tax 725 00:34:41,962 --> 00:34:43,780 cuts that raise money. 726 00:34:43,780 --> 00:34:45,951 PROFESSOR: Well, some claim you do. 727 00:34:45,951 --> 00:34:46,408 You don't actually. 728 00:34:46,408 --> 00:34:48,693 But some claim you have tax cuts that raise money. 729 00:34:48,693 --> 00:34:51,810 That's because they think the elasticity is very large. 730 00:34:51,810 --> 00:34:54,469 If the elasticity is large enough, a tax 731 00:34:54,469 --> 00:34:56,760 cut can raise money. 732 00:34:56,760 --> 00:35:00,920 So, basically, that's all about that some people think 733 00:35:00,920 --> 00:35:02,660 that elasticities are large enough that tax 734 00:35:02,660 --> 00:35:03,180 cuts can raise money. 735 00:35:03,180 --> 00:35:04,530 Those people are wrong. 736 00:35:04,530 --> 00:35:07,780 But that's what they claim. 737 00:35:07,780 --> 00:35:08,709 Yeah. 738 00:35:08,709 --> 00:35:09,959 AUDIENCE: [INAUDIBLE PHRASE]. 739 00:35:12,381 --> 00:35:12,840 PROFESSOR: Yes. 740 00:35:12,840 --> 00:35:13,660 Excellent point. 741 00:35:13,660 --> 00:35:16,390 You'll go through that in section on Friday. 742 00:35:16,390 --> 00:35:18,890 So what I've done is I've done an example of a constant 743 00:35:18,890 --> 00:35:20,020 elasticity curve. 744 00:35:20,020 --> 00:35:22,160 Actually, I've done something here which is logically 745 00:35:22,160 --> 00:35:23,900 inconsistent. 746 00:35:23,900 --> 00:35:26,160 This curve is linear which means it can't be constant 747 00:35:26,160 --> 00:35:27,570 elasticity. 748 00:35:27,570 --> 00:35:31,620 If it's constant elasticity, it would have to curve. 749 00:35:31,620 --> 00:35:36,100 So what I've estimated here is a local elasticity. 750 00:35:36,100 --> 00:35:38,200 I have estimated the elasticity 751 00:35:38,200 --> 00:35:40,550 around that price change. 752 00:35:40,550 --> 00:35:43,590 But the elasticity, if this curve is true, would be 753 00:35:43,590 --> 00:35:46,850 different at different points on this curve. 754 00:35:46,850 --> 00:35:49,020 If the elasticity is going to be constant all over the 755 00:35:49,020 --> 00:35:50,500 curve, and you're going to do a constant elasticity of 756 00:35:50,500 --> 00:35:54,320 demand, that's going to be a curve that bends, 757 00:35:54,320 --> 00:35:55,340 not a linear curve. 758 00:35:55,340 --> 00:35:57,270 So a linear demand curve is not constant 759 00:35:57,270 --> 00:35:58,390 elasticity of demand. 760 00:35:58,390 --> 00:36:00,910 We will typically ignore that issue and focus on local 761 00:36:00,910 --> 00:36:02,440 elasticities. 762 00:36:02,440 --> 00:36:03,440 But that is an important issue. 763 00:36:03,440 --> 00:36:05,030 We'll discuss that in section on Friday, the difference 764 00:36:05,030 --> 00:36:07,590 between constant elasticity of demand curves and linear 765 00:36:07,590 --> 00:36:09,020 demand curves. 766 00:36:09,020 --> 00:36:11,120 But, typically, we're think about local changes. 767 00:36:11,120 --> 00:36:13,650 So if it's local enough, it doesn't really matter. 768 00:36:13,650 --> 00:36:16,120 But, for a broad change, it will matter what the shape of 769 00:36:16,120 --> 00:36:16,940 the curve is. 770 00:36:16,940 --> 00:36:17,930 Good point. 771 00:36:17,930 --> 00:36:19,040 Other questions? 772 00:36:19,040 --> 00:36:20,390 OK. 773 00:36:20,390 --> 00:36:24,160 Let me then turn to another problem we face 774 00:36:24,160 --> 00:36:26,410 in empirical economics. 775 00:36:26,410 --> 00:36:27,850 So this is an example of a problem we're facing in 776 00:36:27,850 --> 00:36:29,680 empirical economics. 777 00:36:29,680 --> 00:36:31,640 Let me turn to an example of another problem we face in 778 00:36:31,640 --> 00:36:34,700 empirical economics estimating elasticities. 779 00:36:34,700 --> 00:36:38,950 It is that individuals often choose the price they face. 780 00:36:41,610 --> 00:36:45,730 Individuals, typically, often don't just face a price that's 781 00:36:45,730 --> 00:36:46,760 given to them. 782 00:36:46,760 --> 00:36:48,100 And then you can say, OK, they're given a price, and we 783 00:36:48,100 --> 00:36:48,700 see how they respond. 784 00:36:48,700 --> 00:36:50,870 They often choose the price they face. 785 00:36:50,870 --> 00:36:53,960 Let me explain what I mean by that. 786 00:36:53,960 --> 00:36:56,380 A classic example of an elasticity that matters a lot 787 00:36:56,380 --> 00:36:59,940 for policies is the elasticity of demand for medical care, 788 00:36:59,940 --> 00:37:02,250 the elasticity of demand for medical care. 789 00:37:02,250 --> 00:37:05,910 That is how much less medical care will you use if you have 790 00:37:05,910 --> 00:37:07,350 to pay for it? 791 00:37:07,350 --> 00:37:10,580 So, for example, most of us have insurance through MIT or 792 00:37:10,580 --> 00:37:12,090 maybe through our parents. 793 00:37:12,090 --> 00:37:14,230 And the way health insurance works is you pay a certain 794 00:37:14,230 --> 00:37:18,270 amount per month or your parents do, and, in return, 795 00:37:18,270 --> 00:37:21,630 that health insurance covers the cost of your medical care, 796 00:37:21,630 --> 00:37:23,020 most of it. 797 00:37:23,020 --> 00:37:26,530 But, typically, you have to pay some of it. 798 00:37:26,530 --> 00:37:29,080 So how many people have gone to the doctor in 799 00:37:29,080 --> 00:37:31,660 the last six months? 800 00:37:31,660 --> 00:37:32,760 Did you have to pay something? 801 00:37:32,760 --> 00:37:33,754 How much did you pay? 802 00:37:33,754 --> 00:37:35,245 Did you pay a copayment? 803 00:37:35,245 --> 00:37:35,742 No? 804 00:37:35,742 --> 00:37:36,736 None of you? 805 00:37:36,736 --> 00:37:37,233 Yeah. 806 00:37:37,233 --> 00:37:38,227 How much did you pay? 807 00:37:38,227 --> 00:37:39,930 AUDIENCE: I think like $20. 808 00:37:39,930 --> 00:37:42,410 PROFESSOR: $20, $10, $5, that's what's called the 809 00:37:42,410 --> 00:37:44,270 copayment, or $0. 810 00:37:44,270 --> 00:37:47,760 Most insurance these days has what's called copayments. 811 00:37:47,760 --> 00:37:51,230 A copayment is what you pay when you go to the doctor. 812 00:37:51,230 --> 00:37:52,586 Insurance picks up the rest. You don't know. 813 00:37:52,586 --> 00:37:54,420 You didn't know how much the whole doctor visit cost. You 814 00:37:54,420 --> 00:37:55,800 just went, you gave them your card. 815 00:37:55,800 --> 00:37:57,025 They said your copayment is $20. 816 00:37:57,025 --> 00:37:57,450 You gave them $20. 817 00:37:57,450 --> 00:37:57,950 You don't know. 818 00:37:57,950 --> 00:38:00,510 The visit might have cost $100, $200, $500, $1,000. 819 00:38:00,510 --> 00:38:01,460 You don't know. 820 00:38:01,460 --> 00:38:05,950 Your insurer picks up the rest. You pay the copayment. 821 00:38:05,950 --> 00:38:11,040 Copayments are rapidly on the rise in health insurance. 822 00:38:11,040 --> 00:38:12,740 There's a rapid rise in copayments. 823 00:38:12,740 --> 00:38:15,860 Increasingly, insurers are saying, look, health care 824 00:38:15,860 --> 00:38:17,700 costs are out of control. 825 00:38:17,700 --> 00:38:20,990 One way we're going to combat them is by making people bear 826 00:38:20,990 --> 00:38:23,700 more of the cost that they use. 827 00:38:23,700 --> 00:38:25,020 I could go on forever about how I'm a health care 828 00:38:25,020 --> 00:38:26,300 economist. I could go on about health care forever. 829 00:38:26,300 --> 00:38:33,120 But just to fix ideas on why this is an issue, in 1950, the 830 00:38:33,120 --> 00:38:37,380 US economy spent 5% of our gross domestic product, 5% of 831 00:38:37,380 --> 00:38:40,100 our size of the economy went to health care. 832 00:38:40,100 --> 00:38:42,570 Today it's 17%. 833 00:38:42,570 --> 00:38:45,930 By 2075, it's projected to be 40%. 834 00:38:45,930 --> 00:38:48,960 That is of every dollar that's made in America, $0.40 will go 835 00:38:48,960 --> 00:38:50,290 to medical care. 836 00:38:50,290 --> 00:38:53,210 By 100 years later, it's about 100%. 837 00:38:53,210 --> 00:38:55,130 Literally, if we do nothing, the entire economy will be 838 00:38:55,130 --> 00:38:56,270 health care. 839 00:38:56,270 --> 00:38:57,770 Obviously, that can't happen. 840 00:38:57,770 --> 00:38:58,970 We've got to deal with this. 841 00:38:58,970 --> 00:39:01,700 And one way that insurers and some policy makers are saying 842 00:39:01,700 --> 00:39:03,840 we need to deal with this is we need to make consumers bear 843 00:39:03,840 --> 00:39:06,260 more of the costs of their medical care. 844 00:39:06,260 --> 00:39:08,220 We need to make consumers pay more when they go to the 845 00:39:08,220 --> 00:39:10,780 doctor, so that they understand the consequences of 846 00:39:10,780 --> 00:39:12,870 their decision. 847 00:39:12,870 --> 00:39:14,980 Well, if we're going to do that, a key question we need 848 00:39:14,980 --> 00:39:17,920 to know is well, does it affect their behavior? 849 00:39:17,920 --> 00:39:21,300 If we make consumers pay more, and it doesn't at all affect 850 00:39:21,300 --> 00:39:24,050 their demand for medical care-- it's just a tax on 851 00:39:24,050 --> 00:39:25,280 them, essentially-- 852 00:39:25,280 --> 00:39:27,220 then that's different than if it causes them to use less 853 00:39:27,220 --> 00:39:28,365 medical care. 854 00:39:28,365 --> 00:39:29,770 It may be good, may be bad. 855 00:39:29,770 --> 00:39:31,390 We'll come back to that. 856 00:39:31,390 --> 00:39:34,850 But the key empirical question is what is the elasticity of 857 00:39:34,850 --> 00:39:36,420 demand for medical care? 858 00:39:36,420 --> 00:39:40,840 If you pay $20 and you pay $0, how much less like are you to 859 00:39:40,840 --> 00:39:44,730 use the doctor when you pay $20 versus when you pay $0. 860 00:39:44,730 --> 00:39:46,960 Well, we can all introspect this and think about it. 861 00:39:46,960 --> 00:39:50,570 But, in fact, to answer this we have to go to the data and 862 00:39:50,570 --> 00:39:53,170 ask, well, what's the difference? 863 00:39:53,170 --> 00:39:55,650 So people, for many years, went to the data. 864 00:39:55,650 --> 00:39:58,030 And they said, look, there's all sorts of differences out 865 00:39:58,030 --> 00:39:59,990 there across people and what they pay for their copayments. 866 00:39:59,990 --> 00:40:01,725 Some people have insurance where they pay nothing, some 867 00:40:01,725 --> 00:40:02,810 where they have $20. 868 00:40:02,810 --> 00:40:05,710 Some people have what they call high-deductible plans. 869 00:40:05,710 --> 00:40:07,900 A deductible plan is where you pay the full cost of your 870 00:40:07,900 --> 00:40:09,950 visits until you reach some limit. 871 00:40:09,950 --> 00:40:13,420 So a $2000 deductible plan will be one where you pay all 872 00:40:13,420 --> 00:40:16,790 of your medical costs until you've spent $2,000. 873 00:40:16,790 --> 00:40:18,050 It's a big copayment. 874 00:40:18,050 --> 00:40:19,780 So we look across those people, and people did. 875 00:40:19,780 --> 00:40:23,380 And they found, look, the people that have plans where 876 00:40:23,380 --> 00:40:26,670 they spend more for health care, where they have a high 877 00:40:26,670 --> 00:40:30,380 copayment, use a lot less health care than where they 878 00:40:30,380 --> 00:40:32,270 don't have to spend anything. 879 00:40:32,270 --> 00:40:36,750 The elasticity of demand looks very, very high. 880 00:40:36,750 --> 00:40:39,600 What is wrong with those studies? 881 00:40:39,600 --> 00:40:41,700 What is wrong with the conclusion those people drew? 882 00:40:41,700 --> 00:40:44,790 They drew it by comparing people who had plans where 883 00:40:44,790 --> 00:40:49,050 they paid a lot to go to the doctor, and therefore use a 884 00:40:49,050 --> 00:40:51,635 lot less care to people who didn't pay anything when they 885 00:40:51,635 --> 00:40:53,911 went to the doctor and used a lot more care. 886 00:40:53,911 --> 00:40:57,210 I pick the $20 person, because I picked on you already. 887 00:40:57,210 --> 00:41:00,460 AUDIENCE: Probably they chose to have a high-deductible 888 00:41:00,460 --> 00:41:04,120 plan, because they don't often go to the doctor already. 889 00:41:04,120 --> 00:41:07,810 PROFESSOR: The rational choice, if you're young and 890 00:41:07,810 --> 00:41:10,830 healthy, for almost everyone in this room, is going to be a 891 00:41:10,830 --> 00:41:13,810 very high-deductible, high copayment plan. 892 00:41:13,810 --> 00:41:16,930 Because it will cost you less money, because the insurer is 893 00:41:16,930 --> 00:41:18,045 shifting the money to you. 894 00:41:18,045 --> 00:41:19,300 But you don't use the doctor anyway. 895 00:41:19,300 --> 00:41:21,510 So who cares? 896 00:41:21,510 --> 00:41:23,820 So the healthier people are going to choose the plans 897 00:41:23,820 --> 00:41:24,800 where they pay more. 898 00:41:24,800 --> 00:41:27,375 So, of course, you're going to find in the plans where people 899 00:41:27,375 --> 00:41:32,780 pay more they use less medical care. 900 00:41:32,780 --> 00:41:35,300 But is it because they're paying more, or is it because 901 00:41:35,300 --> 00:41:37,490 healthy guys choose those plans? 902 00:41:37,490 --> 00:41:39,820 It's causation versus correlation. 903 00:41:39,820 --> 00:41:42,120 We don't know. 904 00:41:42,120 --> 00:41:44,400 Well, how can we figure that out? 905 00:41:44,400 --> 00:41:48,460 Well, if we were doctors, what we'd do-- real doctors, not a 906 00:41:48,460 --> 00:41:51,760 doctor like me, a real doctor, a medical doctor-- 907 00:41:51,760 --> 00:41:55,980 what we'd do is we'd run a randomized trial. 908 00:41:55,980 --> 00:42:00,320 So if doctors want to figure out whether a drug works or 909 00:42:00,320 --> 00:42:04,810 not, they don't just look at guys who take the drug versus 910 00:42:04,810 --> 00:42:06,790 guys who don't. 911 00:42:06,790 --> 00:42:07,820 They run a randomized trial. 912 00:42:07,820 --> 00:42:11,130 They randomly assign some people to take the drug and 913 00:42:11,130 --> 00:42:13,431 some people not. 914 00:42:13,431 --> 00:42:16,840 Now, when you run a randomized trial, by definition, you get 915 00:42:16,840 --> 00:42:19,500 a causal effect. 916 00:42:19,500 --> 00:42:21,510 Well, this room isn't quite big enough. 917 00:42:21,510 --> 00:42:23,230 We all know the law of large numbers. 918 00:42:23,230 --> 00:42:24,990 But imagine there were four times as many people in this 919 00:42:24,990 --> 00:42:27,000 room or five times as many people in this room. 920 00:42:27,000 --> 00:42:27,620 OK? 921 00:42:27,620 --> 00:42:28,990 And I had you come up to the front. 922 00:42:28,990 --> 00:42:31,510 I flipped a coin and said half of you are going to take the 923 00:42:31,510 --> 00:42:34,470 drug, and half of you are not, randomly by 924 00:42:34,470 --> 00:42:36,260 the flip of a coin. 925 00:42:36,260 --> 00:42:39,400 Then, by definition, any statistically noticeable 926 00:42:39,400 --> 00:42:42,320 differences I get between the group the takes the drug and 927 00:42:42,320 --> 00:42:44,740 the group that doesn't is caused by the drug. 928 00:42:44,740 --> 00:42:45,860 And how do I know that? 929 00:42:45,860 --> 00:42:48,010 Because I know the groups are otherwise identical by the law 930 00:42:48,010 --> 00:42:49,090 of large numbers. 931 00:42:49,090 --> 00:42:51,140 By the law of large numbers, I know that as long as I have 932 00:42:51,140 --> 00:42:54,050 enough people, they're identical. 933 00:42:54,050 --> 00:42:55,800 So if the only difference between them is that one's 934 00:42:55,800 --> 00:42:59,390 taking the drug and one's not, that's a randomized trial. 935 00:42:59,390 --> 00:43:01,540 That would be how I could solve the causation versus 936 00:43:01,540 --> 00:43:03,820 correlation problem. 937 00:43:03,820 --> 00:43:07,570 In medicine, thousands of randomized trials every day 938 00:43:07,570 --> 00:43:09,380 are being run. 939 00:43:09,380 --> 00:43:11,920 In fact, the FDA, before it will approve a drug, will 940 00:43:11,920 --> 00:43:14,040 typically require a randomized trial. 941 00:43:14,040 --> 00:43:15,820 Well, in the social sciences, it's harder to 942 00:43:15,820 --> 00:43:16,770 run randomized trials. 943 00:43:16,770 --> 00:43:20,220 Because we're actually trying to understand things like 944 00:43:20,220 --> 00:43:22,190 people's demand for medical care, not whether a 945 00:43:22,190 --> 00:43:23,390 drug works or not. 946 00:43:23,390 --> 00:43:26,430 But, in fact, one of the most famous social randomized 947 00:43:26,430 --> 00:43:28,680 trials in history was called the RAND Health Insurance 948 00:43:28,680 --> 00:43:31,820 Experiment run in the 1970s. 949 00:43:31,820 --> 00:43:33,450 This is where some innovative health economists who 950 00:43:33,450 --> 00:43:35,720 understood this problem that we laid out about the fact 951 00:43:35,720 --> 00:43:37,330 that you can't just compare more or less generous health 952 00:43:37,330 --> 00:43:40,350 insurance policies, actually randomized health insurance 953 00:43:40,350 --> 00:43:42,310 policies across people. 954 00:43:42,310 --> 00:43:46,460 They recruited volunteers, and they literally said, we're 955 00:43:46,460 --> 00:43:47,700 going to randomize. 956 00:43:47,700 --> 00:43:49,580 Some people are going to have policies where the health care 957 00:43:49,580 --> 00:43:54,580 is free, and some people are going to have policies where 958 00:43:54,580 --> 00:43:58,840 they have to pay, essentially, all the costs of health care. 959 00:43:58,840 --> 00:44:01,370 So they, essentially, randomized across these 960 00:44:01,370 --> 00:44:02,530 different groups. 961 00:44:02,530 --> 00:44:05,510 And, therefore, they can assess what the price 962 00:44:05,510 --> 00:44:06,360 elasticity was. 963 00:44:06,360 --> 00:44:09,370 Because they knew the price difference between groups. 964 00:44:09,370 --> 00:44:10,370 For one, the price was zero. 965 00:44:10,370 --> 00:44:11,580 For one, the price was one. 966 00:44:11,580 --> 00:44:14,710 They actually had a range of prices they varied it across. 967 00:44:14,710 --> 00:44:16,480 They could look at the quantity response, and they 968 00:44:16,480 --> 00:44:18,280 knew that was a quantity response to the price, because 969 00:44:18,280 --> 00:44:19,810 people weren't choosing their prices. 970 00:44:19,810 --> 00:44:22,450 The prices were being assigned to them. 971 00:44:22,450 --> 00:44:23,680 What did they find? 972 00:44:23,680 --> 00:44:26,940 Well, they found that medical demand is elastic, although 973 00:44:26,940 --> 00:44:28,690 not as elastic as the previous study. 974 00:44:28,690 --> 00:44:30,375 It's somewhat elastic. 975 00:44:30,375 --> 00:44:33,200 It's not as elastic as the previous studies found. 976 00:44:33,200 --> 00:44:35,260 They found that the elasticity of demand for medical care is 977 00:44:35,260 --> 00:44:37,950 around minus 0.2. 978 00:44:37,950 --> 00:44:41,710 So when the price goes up, people use less medical care 979 00:44:41,710 --> 00:44:45,540 but not that much less. 980 00:44:45,540 --> 00:44:46,790 Now, let's be clear. 981 00:44:46,790 --> 00:44:47,770 Remember what elasticity is. 982 00:44:47,770 --> 00:44:51,490 That delta Q over Q. The same study showed that if you take 983 00:44:51,490 --> 00:44:55,160 someone who paid nothing and make them pay almost 984 00:44:55,160 --> 00:45:00,400 everything, their utilization of medical care falls by 45%. 985 00:45:00,400 --> 00:45:01,830 That's consistent with that small elasticity. 986 00:45:01,830 --> 00:45:06,280 Because that's a huge delta P, percent delta P. So, 987 00:45:06,280 --> 00:45:10,130 basically, that's comes to the question about local versus 988 00:45:10,130 --> 00:45:11,500 global elasticities. 989 00:45:11,500 --> 00:45:14,060 So it's not saying that prices don't matter. 990 00:45:14,060 --> 00:45:18,270 But it's not a very, very elastically demanded good. 991 00:45:18,270 --> 00:45:19,710 So that's how they measure that price of 992 00:45:19,710 --> 00:45:20,960 elasticity of demand. 993 00:45:22,990 --> 00:45:27,620 That experiment, which was run over 35 years ago now, that 994 00:45:27,620 --> 00:45:30,930 result drives much of what we do in health policy. 995 00:45:30,930 --> 00:45:34,370 So a lot of the estimates that we saw for the recently passed 996 00:45:34,370 --> 00:45:37,490 health reform bill derived from how 997 00:45:37,490 --> 00:45:38,080 do we get that estimate. 998 00:45:38,080 --> 00:45:39,650 We'll have to figure out how people are going to respond 999 00:45:39,650 --> 00:45:42,250 with their medical care when we give them health insurance. 1000 00:45:42,250 --> 00:45:45,260 The recently passed health care bill just gave 32 million 1001 00:45:45,260 --> 00:45:46,420 people health insurance. 1002 00:45:46,420 --> 00:45:48,170 Well, how are they going to respond to 1003 00:45:48,170 --> 00:45:50,660 having health insurance? 1004 00:45:50,660 --> 00:45:52,450 We go back to the RAND estimates and say, well, we 1005 00:45:52,450 --> 00:45:53,510 have this elasticity of demand. 1006 00:45:53,510 --> 00:45:54,480 We know what we're doing to the price. 1007 00:45:54,480 --> 00:45:58,840 We figure out how much medical care is going to go up. 1008 00:45:58,840 --> 00:46:00,370 But here's the other thing. 1009 00:46:00,370 --> 00:46:01,420 Here's the question in the lecture that 1010 00:46:01,420 --> 00:46:02,640 that we'll close with. 1011 00:46:02,640 --> 00:46:06,030 Is that a good thing or a bad thing that medical care fell 1012 00:46:06,030 --> 00:46:07,570 when the price went up? 1013 00:46:07,570 --> 00:46:09,990 And how would we tell whether it's a good 1014 00:46:09,990 --> 00:46:11,210 thing or a bad thing? 1015 00:46:11,210 --> 00:46:12,810 So we know when we raise the price, people use 1016 00:46:12,810 --> 00:46:13,960 less medical care. 1017 00:46:13,960 --> 00:46:16,920 How can we tell if that's a good thing or a bad thing? 1018 00:46:16,920 --> 00:46:18,890 In the same experiment, how could we tell? 1019 00:46:18,890 --> 00:46:20,370 What could we do? 1020 00:46:20,370 --> 00:46:21,100 Yeah. 1021 00:46:21,100 --> 00:46:23,590 AUDIENCE: Maybe you'll get death rates or like-- 1022 00:46:23,590 --> 00:46:25,190 PROFESSOR: You look at their health. 1023 00:46:25,190 --> 00:46:27,965 You say, look, the same trial can 1024 00:46:27,965 --> 00:46:29,210 answer a different question. 1025 00:46:29,210 --> 00:46:30,790 We know that when you charge someone for health 1026 00:46:30,790 --> 00:46:31,620 care, they use less. 1027 00:46:31,620 --> 00:46:33,345 Well, are they sicker? 1028 00:46:33,345 --> 00:46:36,340 The answer, not at all. 1029 00:46:36,340 --> 00:46:40,450 People use less health and were no sicker. 1030 00:46:40,450 --> 00:46:41,200 Why? 1031 00:46:41,200 --> 00:46:45,250 Because we waste a huge amount of health care in the US. 1032 00:46:45,250 --> 00:46:46,710 A huge amount of health care is wasted. 1033 00:46:46,710 --> 00:46:50,190 So, in fact, we could cut back quite a lot on health care, 1034 00:46:50,190 --> 00:46:51,500 and we'd be no sicker. 1035 00:46:51,500 --> 00:46:53,160 And that's what the RAND experiment showed, that we can 1036 00:46:53,160 --> 00:46:56,080 charge people to use medical providers. 1037 00:46:56,080 --> 00:46:58,820 And they'll use less medical care, and they won't be sicker 1038 00:46:58,820 --> 00:46:59,580 as a result. 1039 00:46:59,580 --> 00:47:01,520 Which suggests that, actually, as we try to think about 1040 00:47:01,520 --> 00:47:03,770 getting our health care costs under control in America, 1041 00:47:03,770 --> 00:47:06,570 making people pay something to go to the doctor is not a 1042 00:47:06,570 --> 00:47:07,930 crazy thing to be thinking about. 1043 00:47:07,930 --> 00:47:08,425 How much? 1044 00:47:08,425 --> 00:47:11,220 Well that depends on efficiency versus equity. 1045 00:47:11,220 --> 00:47:14,860 We can't make someone who has no income pay $1,000 to go to 1046 00:47:14,860 --> 00:47:15,650 the doctor. 1047 00:47:15,650 --> 00:47:17,590 That, clearly, is a mistake. 1048 00:47:17,590 --> 00:47:20,840 But we can take a rich guy like me and make me pay $50 to 1049 00:47:20,840 --> 00:47:22,840 go to the doctor. 1050 00:47:22,840 --> 00:47:25,260 There's no reason not to do that. 1051 00:47:25,260 --> 00:47:28,140 So, basically, that's a lesson of how you can use elasticity 1052 00:47:28,140 --> 00:47:29,960 of demand to help inform the kind of 1053 00:47:29,960 --> 00:47:31,340 policies we need to make. 1054 00:47:31,340 --> 00:47:32,830 OK, let me stop there. 1055 00:47:32,830 --> 00:47:35,600 By the way, if you at all find this stuff interesting, and 1056 00:47:35,600 --> 00:47:36,930 you haven't yet read Freakonomics-- 1057 00:47:36,930 --> 00:47:38,740 how many of you have read Freakonomics? 1058 00:47:38,740 --> 00:47:39,980 That's amazing. 1059 00:47:39,980 --> 00:47:40,370 OK. 1060 00:47:40,370 --> 00:47:41,670 If you haven't read Freakonomics, you should. 1061 00:47:41,670 --> 00:47:42,450 It's a great book. 1062 00:47:42,450 --> 00:47:44,600 If you're lazy, the movie is coming out. 1063 00:47:44,600 --> 00:47:47,230 And Freakonomics the movie is premiering on 1064 00:47:47,230 --> 00:47:50,090 Friday the 30th at LSC. 1065 00:47:50,090 --> 00:47:52,200 So if you're interested in learning more about empirical 1066 00:47:52,200 --> 00:47:54,860 tools in economics, you can watch Freakonomics the movie 1067 00:47:54,860 --> 00:47:56,880 on Friday the 30th.