WEBVTT 00:00:01.070 --> 00:00:03.410 The following content is provided under a Creative 00:00:03.410 --> 00:00:04.830 Commons license. 00:00:04.830 --> 00:00:07.040 Your support will help MIT Open CourseWare 00:00:07.040 --> 00:00:11.130 continue to offer high quality educational resources for free. 00:00:11.130 --> 00:00:13.670 To make a donation or to view additional materials 00:00:13.670 --> 00:00:17.630 from hundreds of MIT courses, visit MIT OpenCourseWare 00:00:17.630 --> 00:00:18.510 at ocw.mit.edu. 00:00:22.400 --> 00:00:25.200 GABRIEL SANCHEZ-MARTINEZ: So let's get started today. 00:00:25.200 --> 00:00:28.820 Again, the topic is fare policy, structure, and technology. 00:00:28.820 --> 00:00:32.420 So we'll cover policy objectives, the issues 00:00:32.420 --> 00:00:35.140 that agencies face, fare structure, 00:00:35.140 --> 00:00:39.150 demand response to fare changes and fare technology. 00:00:39.150 --> 00:00:41.810 So what are the objectives of fare policy? 00:00:45.860 --> 00:00:49.850 There are multiple objectives, and they don't necessarily 00:00:49.850 --> 00:00:50.930 support each other. 00:00:50.930 --> 00:00:52.680 In fact, they often go against each other. 00:00:52.680 --> 00:00:56.340 So the number one objective is to fund operations. 00:00:56.340 --> 00:01:01.250 That's the reason there's a user fee for service. 00:01:01.250 --> 00:01:02.660 And I say here at least partially 00:01:02.660 --> 00:01:05.840 because for the most part, the revenues obtained 00:01:05.840 --> 00:01:09.290 by fare collection do not cover the full operational costs, 00:01:09.290 --> 00:01:11.280 let alone capital costs. 00:01:11.280 --> 00:01:14.990 So here we have some examples of fare recovery ratios 00:01:14.990 --> 00:01:18.965 based on the 2014 National Transit Database. 00:01:18.965 --> 00:01:20.660 Now, you might remember that we talked 00:01:20.660 --> 00:01:22.310 about the National Transit Database 00:01:22.310 --> 00:01:23.870 earlier in this course. 00:01:23.870 --> 00:01:26.370 And here is a breakdown. 00:01:26.370 --> 00:01:28.940 I looked at all heavy rail and I saw 00:01:28.940 --> 00:01:31.770 a range of between 0.14 0.78. 00:01:31.770 --> 00:01:35.525 So between 14% and 78% of operational expenses 00:01:35.525 --> 00:01:39.260 are covered by fare revenue. 00:01:39.260 --> 00:01:42.500 The highest one here is in California. 00:01:42.500 --> 00:01:47.327 For light rail a little lower and for bus all over the place. 00:01:47.327 --> 00:01:48.410 Here we have the averages. 00:01:48.410 --> 00:01:52.880 These tend to be higher for rail and lower for bus. 00:01:52.880 --> 00:01:55.130 So why is that? 00:01:55.130 --> 00:01:57.044 Why would that be the case? 00:01:57.044 --> 00:01:58.460 AUDIENCE: On an operational level, 00:01:58.460 --> 00:02:02.270 one rail driver is carrying a lot more passengers 00:02:02.270 --> 00:02:03.919 than one bus driver. 00:02:03.919 --> 00:02:05.210 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:02:05.210 --> 00:02:12.684 Although you can fill up a bus, right? 00:02:12.684 --> 00:02:15.100 AUDIENCE: You can fill up a bus [INAUDIBLE] train carrying 00:02:15.100 --> 00:02:15.770 500 passengers. 00:02:15.770 --> 00:02:16.160 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:02:16.160 --> 00:02:17.270 Yeah. 00:02:17.270 --> 00:02:17.820 Yeah. 00:02:17.820 --> 00:02:19.560 Yeah. 00:02:19.560 --> 00:02:25.020 So to sort of go back to my rebuttal, 00:02:25.020 --> 00:02:29.010 there are some buzzers here with fare recovery ratios over one. 00:02:29.010 --> 00:02:34.370 So it's possible to have a bus service that returns profit 00:02:34.370 --> 00:02:35.100 even. 00:02:35.100 --> 00:02:40.020 So why is it that system-wide, bus services 00:02:40.020 --> 00:02:42.960 tend to have lower fare recovery ratios? 00:02:42.960 --> 00:02:45.390 Because you have a good point, the productivity of rail 00:02:45.390 --> 00:02:46.710 can be higher. 00:02:46.710 --> 00:02:49.680 Of course, if rail is underutilized, 00:02:49.680 --> 00:02:51.660 then you still have that issue. 00:02:51.660 --> 00:02:53.040 Any other ideas? 00:02:53.040 --> 00:02:54.324 [INAUDIBLE] 00:02:54.324 --> 00:02:58.140 AUDIENCE: Buses [INAUDIBLE] who can have the most service 00:02:58.140 --> 00:03:01.587 and cover the most low ridership routes? 00:03:01.587 --> 00:03:02.920 GABRIEL SANCHEZ-MARTINEZ: Right. 00:03:02.920 --> 00:03:05.110 So there's an agency and you system-wide 00:03:05.110 --> 00:03:07.360 have to cover your territory. 00:03:07.360 --> 00:03:11.590 That will include buses that are not very heavily utilized. 00:03:11.590 --> 00:03:13.521 Also, you're providing service late 00:03:13.521 --> 00:03:14.770 at night and all those things. 00:03:14.770 --> 00:03:19.540 So there are many cases of buses not being very well utilized, 00:03:19.540 --> 00:03:22.180 and that brings down the fare recovery ratio. 00:03:22.180 --> 00:03:24.650 I did not include here demand responsive. 00:03:24.650 --> 00:03:30.250 Demand responsive is usually around 1% to 2%. 00:03:30.250 --> 00:03:32.680 So I'm talking about what here in Boston 00:03:32.680 --> 00:03:38.080 is called the ride for disabled passengers. 00:03:38.080 --> 00:03:43.360 And it can cost around $20 per ride, 00:03:43.360 --> 00:03:44.920 and people pay a lot less. 00:03:44.920 --> 00:03:48.897 So I think $20 is low. 00:03:48.897 --> 00:03:49.730 I think it's higher. 00:03:49.730 --> 00:03:50.450 Sorry. 00:03:50.450 --> 00:03:53.060 I don't remember the number, but it's closer to $30 or $40, 00:03:53.060 --> 00:03:54.990 I think. 00:03:54.990 --> 00:03:59.630 So the fare recovery ratio for demand responsive transit 00:03:59.630 --> 00:04:00.870 is much lower. 00:04:00.870 --> 00:04:01.370 OK. 00:04:01.370 --> 00:04:02.840 So other goals. 00:04:02.840 --> 00:04:05.640 Keep transit affordable and promote social equity. 00:04:05.640 --> 00:04:07.940 So here we have the first conflicts, right? 00:04:07.940 --> 00:04:11.030 The first objective would mean that you want to raise fares, 00:04:11.030 --> 00:04:14.000 and if you raise fares then you get more fare revenue. 00:04:14.000 --> 00:04:15.500 But there's another objective, which 00:04:15.500 --> 00:04:18.410 is to keep transit affordable and for social equity, 00:04:18.410 --> 00:04:21.019 and that would indicate that you want to keep fares low. 00:04:21.019 --> 00:04:22.850 Right? 00:04:22.850 --> 00:04:26.780 So also, supporting growth for demand of transit 00:04:26.780 --> 00:04:29.630 also keeps fares low. 00:04:29.630 --> 00:04:32.180 Make the fare structure easy to communicate. 00:04:32.180 --> 00:04:35.060 We'll talk about different fare structures and market 00:04:35.060 --> 00:04:37.050 segmentation later in this lecture. 00:04:37.050 --> 00:04:40.880 But the point is that when you do market segmentation, 00:04:40.880 --> 00:04:45.560 you can get more revenue out of your fare system. 00:04:45.560 --> 00:04:47.420 When you want to keep things simple, 00:04:47.420 --> 00:04:50.300 then you're probably going to suggest going for a flat fare 00:04:50.300 --> 00:04:53.900 structure, where people that have much higher willingness 00:04:53.900 --> 00:04:56.120 to pay are still paying low fares. 00:04:56.120 --> 00:04:59.240 So that objective of keeping the fare 00:04:59.240 --> 00:05:02.090 structure easy to communicate and simple 00:05:02.090 --> 00:05:04.670 will reduce fare earnings. 00:05:04.670 --> 00:05:07.170 And finally, to reduce fare system costs. 00:05:07.170 --> 00:05:10.640 So there are costs to fare collection 00:05:10.640 --> 00:05:12.650 and to the fare system as a whole. 00:05:12.650 --> 00:05:13.950 Fare collection is one of them. 00:05:13.950 --> 00:05:18.440 So obviously, the capital cost of installing the fare boxes 00:05:18.440 --> 00:05:21.770 and maintaining the whole AFC system. 00:05:21.770 --> 00:05:24.800 There is a cost for people who are maintaining those fare 00:05:24.800 --> 00:05:29.000 boxes and vaulting cash out of the fare boxes 00:05:29.000 --> 00:05:32.630 and counting the cash and keeping it in a safe place. 00:05:32.630 --> 00:05:33.950 And there's customer service. 00:05:33.950 --> 00:05:37.940 So customer service involves more than the fare system. 00:05:37.940 --> 00:05:39.830 But some aspects of customer service 00:05:39.830 --> 00:05:44.750 involve my CharlieCard doesn't work, can I get a new one? 00:05:44.750 --> 00:05:47.510 Or I already paid for that ride and you 00:05:47.510 --> 00:05:49.260 need to give me a refund. 00:05:49.260 --> 00:05:53.930 So these kinds of inquiries or complaints 00:05:53.930 --> 00:05:56.900 are coming from the fare system. 00:05:56.900 --> 00:05:57.400 OK? 00:05:57.400 --> 00:06:01.060 Questions about these objectives? 00:06:01.060 --> 00:06:01.560 No. 00:06:01.560 --> 00:06:02.360 OK. 00:06:02.360 --> 00:06:05.770 So how does fare policy intersect with other areas? 00:06:05.770 --> 00:06:08.400 It intersects with many areas, and that makes fare policy 00:06:08.400 --> 00:06:12.530 and fare structure a very powerful sort of lever or knob 00:06:12.530 --> 00:06:15.170 for a transit agency. 00:06:15.170 --> 00:06:17.390 So obviously, finance. 00:06:17.390 --> 00:06:19.400 We're talking about funding operational expenses 00:06:19.400 --> 00:06:22.520 from the revenues derived from the fare system. 00:06:22.520 --> 00:06:23.820 And then there's operations. 00:06:23.820 --> 00:06:29.630 So if you have a fare policy that makes people board faster 00:06:29.630 --> 00:06:32.100 because, for example, it's proof of payments 00:06:32.100 --> 00:06:33.920 so there's no validation required 00:06:33.920 --> 00:06:37.790 or you have fare technology that a smart card is 00:06:37.790 --> 00:06:40.820 much faster than paying cash, then 00:06:40.820 --> 00:06:43.340 that means that you'll have faster boarding, shorter dwell 00:06:43.340 --> 00:06:46.430 times, shorter cycle times, higher reliability. 00:06:46.430 --> 00:06:51.800 So you might affect operations through fare technology. 00:06:51.800 --> 00:06:54.470 Some fare structures require fare inspection. 00:06:54.470 --> 00:06:58.070 So proof of payment is one that people board 00:06:58.070 --> 00:07:00.570 without interacting with any fare system, 00:07:00.570 --> 00:07:02.490 so then an inspector has to come in 00:07:02.490 --> 00:07:06.140 and either for all passengers or more likely 00:07:06.140 --> 00:07:09.320 for a subset of passengers, a sample of passengers, request 00:07:09.320 --> 00:07:12.350 proof of having paid the fare. 00:07:12.350 --> 00:07:15.680 And of course, that involves some cost. 00:07:15.680 --> 00:07:17.096 There's maintenance of equipment. 00:07:17.096 --> 00:07:18.470 If equipment breaks down, the bus 00:07:18.470 --> 00:07:22.220 needs to be taken out of service, et cetera. 00:07:22.220 --> 00:07:23.130 Public support. 00:07:23.130 --> 00:07:26.990 So politicians may promise not raising fares. 00:07:26.990 --> 00:07:28.812 That's a popular campaign promise. 00:07:28.812 --> 00:07:30.770 And then of course, that puts a lot of pressure 00:07:30.770 --> 00:07:33.650 on the agency to do more with less. 00:07:33.650 --> 00:07:36.140 And we'll talk more about that when 00:07:36.140 --> 00:07:38.820 we discuss transit finance. 00:07:38.820 --> 00:07:42.230 There's difficulty gathering support to raise fares. 00:07:42.230 --> 00:07:45.190 Usually, we'll improve service first 00:07:45.190 --> 00:07:47.580 and then that justifies raising fares. 00:07:47.580 --> 00:07:48.080 Right? 00:07:48.080 --> 00:07:50.060 I don't want my service to get more expensive 00:07:50.060 --> 00:07:52.956 if it's not very good service. 00:07:52.956 --> 00:07:54.937 AUDIENCE: [INAUDIBLE] 00:07:54.937 --> 00:07:56.270 GABRIEL SANCHEZ-MARTINEZ: Right. 00:07:56.270 --> 00:07:59.270 And essentially, if you have lower service quality, 00:07:59.270 --> 00:08:01.761 then fewer people end up taking it or liking it, 00:08:01.761 --> 00:08:03.260 and those people are not going to be 00:08:03.260 --> 00:08:09.260 the ones rallying for a transit sort of fare hike. 00:08:09.260 --> 00:08:11.627 Then there's also labor's push for higher wages. 00:08:11.627 --> 00:08:13.460 So that's another sort of political pressure 00:08:13.460 --> 00:08:17.090 on the agency that increases operational costs 00:08:17.090 --> 00:08:20.360 and therefore, may require an increase in fares. 00:08:20.360 --> 00:08:24.140 But both a labor dispute and a dispute 00:08:24.140 --> 00:08:26.900 with the passengers on the fare hike 00:08:26.900 --> 00:08:32.210 are the result of unpleasant negotiations or interactions 00:08:32.210 --> 00:08:33.770 in public. 00:08:33.770 --> 00:08:36.907 Administration, so fare technology and fare policy 00:08:36.907 --> 00:08:37.740 and equity analysis. 00:08:37.740 --> 00:08:41.750 If you're considering any change to technology or policy 00:08:41.750 --> 00:08:45.020 or structure, you have to analyze the economic aspects 00:08:45.020 --> 00:08:51.170 of it and you have to prove that you're not 00:08:51.170 --> 00:08:54.850 discriminating against any one group, even unintentionally. 00:08:54.850 --> 00:08:58.430 So disproportionately affecting a disadvantaged community, 00:08:58.430 --> 00:09:05.480 either by race or income, that's called Title VI in the US. 00:09:05.480 --> 00:09:07.850 Then there's another big policy issue 00:09:07.850 --> 00:09:10.160 is revenue sharing across jurisdictions. 00:09:10.160 --> 00:09:14.840 Not so much an issue on the east coast of the US, but certainly 00:09:14.840 --> 00:09:18.590 if you go to the West Coast, you have in the San Francisco Bay 00:09:18.590 --> 00:09:22.050 Area dozens of operators. 00:09:22.050 --> 00:09:24.180 Some of them share the same fare system. 00:09:24.180 --> 00:09:29.570 So how much does each operator pay 00:09:29.570 --> 00:09:32.780 for routes that serve multiple jurisdictions? 00:09:32.780 --> 00:09:36.410 That's one question. 00:09:36.410 --> 00:09:39.760 The other question might be, if we are separate operators 00:09:39.760 --> 00:09:42.710 and we each have a route so we take care of that, 00:09:42.710 --> 00:09:46.430 we still are collecting fares with a joint system. 00:09:46.430 --> 00:09:47.780 Do we trust each other? 00:09:47.780 --> 00:09:51.530 And often, actually, the barriers to integration 00:09:51.530 --> 00:09:55.770 are more about trust and less about technology. 00:09:55.770 --> 00:09:57.904 Marketing is another sort of issue 00:09:57.904 --> 00:09:59.070 the agency has to deal with. 00:09:59.070 --> 00:10:02.640 How much do they market and what is the message? 00:10:02.640 --> 00:10:05.460 And customer service, which we already mentioned. 00:10:05.460 --> 00:10:07.260 It's interesting to note that fares 00:10:07.260 --> 00:10:09.390 are one of the first things that a person 00:10:09.390 --> 00:10:11.901 has to sort of understand about a transit system 00:10:11.901 --> 00:10:12.900 to be able to take them. 00:10:12.900 --> 00:10:15.160 So if you're a tourist and you go to a new city 00:10:15.160 --> 00:10:17.670 that you've never been to before you arrive at the airport 00:10:17.670 --> 00:10:20.430 and you want to take the bus or the train from the airport, 00:10:20.430 --> 00:10:21.804 the first thing you have to do is 00:10:21.804 --> 00:10:24.240 buy a ticket, right, and figure out the fare structure 00:10:24.240 --> 00:10:27.570 and where are the maps and et cetera. 00:10:27.570 --> 00:10:30.840 So this is sort of one of the first issues, 00:10:30.840 --> 00:10:33.450 the first impression that a passenger has about the fare 00:10:33.450 --> 00:10:36.390 system, about the transit agency, something 00:10:36.390 --> 00:10:39.720 related to the fare system. 00:10:39.720 --> 00:10:40.220 OK. 00:10:40.220 --> 00:10:44.800 What issues do agencies face regarding these topics? 00:10:44.800 --> 00:10:46.420 One is fare recovery ratios. 00:10:46.420 --> 00:10:48.390 So we looked at ranges. 00:10:48.390 --> 00:10:50.790 Typically, fare recovery ratio system-wide across modes 00:10:50.790 --> 00:10:52.200 is about 1/3. 00:10:52.200 --> 00:10:54.940 It varies a lot, as we saw. 00:10:54.940 --> 00:10:57.160 It's rare to make a profit system. 00:10:57.160 --> 00:11:00.120 So one of the issues is how much-- 00:11:00.120 --> 00:11:03.030 well, the first question is, how much of a fare recovery ratio 00:11:03.030 --> 00:11:09.720 do we need or do we set a goal to cover operational expenses? 00:11:09.720 --> 00:11:12.930 And is it sort of a strong requirement, 00:11:12.930 --> 00:11:14.420 as it is in some agencies? 00:11:14.420 --> 00:11:18.540 Some agencies have laws that say the fare recovery ratio must 00:11:18.540 --> 00:11:19.950 be at least this much. 00:11:19.950 --> 00:11:24.270 So if expenses go up, automatic fare hike. 00:11:24.270 --> 00:11:28.640 Others, it's more of a guideline or maybe there is no policy. 00:11:28.640 --> 00:11:31.150 So that that's one of the key issues. 00:11:31.150 --> 00:11:33.630 The second one is how often to raise fares. 00:11:33.630 --> 00:11:36.270 I think most agencies are reactive. 00:11:36.270 --> 00:11:38.370 So a reactive approach is one where 00:11:38.370 --> 00:11:41.220 you avoid the headache of a fare hike 00:11:41.220 --> 00:11:46.110 and all the sort of public engagement that goes with it 00:11:46.110 --> 00:11:49.590 and you only do that when you have to so 00:11:49.590 --> 00:11:54.600 you are in a financial situation where you absolutely 00:11:54.600 --> 00:11:59.880 need to raise fare revenues to continue operating. 00:11:59.880 --> 00:12:03.370 So that happens less than once every year 00:12:03.370 --> 00:12:05.280 and it's a big effort. 00:12:05.280 --> 00:12:08.400 Some agencies take the approach of an annual fare 00:12:08.400 --> 00:12:10.196 hike with inflation. 00:12:10.196 --> 00:12:11.820 So it's kind of predictable, it happens 00:12:11.820 --> 00:12:13.380 every year at the same date. 00:12:13.380 --> 00:12:16.590 It tends to make each one easier than one 00:12:16.590 --> 00:12:20.160 done ad hoc because it's more expectable 00:12:20.160 --> 00:12:23.420 and there's less public outrage. 00:12:26.222 --> 00:12:28.180 So gathering and maintaining political support, 00:12:28.180 --> 00:12:30.910 we talked a little bit about that. 00:12:30.910 --> 00:12:35.270 Of course, that is an issue that will play out and sort of it 00:12:35.270 --> 00:12:40.570 will affect how other issues are faced and dealt with. 00:12:40.570 --> 00:12:43.030 One other issue is do we raise base fares 00:12:43.030 --> 00:12:47.680 or do we change the relative cost of passes and discounted 00:12:47.680 --> 00:12:48.980 fare products? 00:12:48.980 --> 00:12:52.300 So if a base fare is $2, you might 00:12:52.300 --> 00:12:57.610 have 50% subsidy for seniors and you 00:12:57.610 --> 00:13:01.570 might have some other discounts for students or other groups. 00:13:01.570 --> 00:13:04.600 You might also have passes like a weekly pass, which 00:13:04.600 --> 00:13:07.210 if you take some number of trips you break even 00:13:07.210 --> 00:13:09.160 and after that, you ride free. 00:13:09.160 --> 00:13:12.880 So you might actually leave the base fare the same 00:13:12.880 --> 00:13:17.350 and change the percentages of subsidy or discounts 00:13:17.350 --> 00:13:20.020 or the break even point of a pass. 00:13:20.020 --> 00:13:21.550 And that can affect fare revenues, 00:13:21.550 --> 00:13:24.010 it can affect ridership, and it can affect people's choice 00:13:24.010 --> 00:13:25.700 of fare product. 00:13:25.700 --> 00:13:32.980 So you might bump people towards letting go of fare passes 00:13:32.980 --> 00:13:37.480 and opting to do pay as you go more often or the other way 00:13:37.480 --> 00:13:38.240 around. 00:13:38.240 --> 00:13:39.910 And that has implications on how people 00:13:39.910 --> 00:13:45.520 think about the transit system, how much it costs for them. 00:13:45.520 --> 00:13:49.330 For a pass holder, of course, once they've paid for it, 00:13:49.330 --> 00:13:52.630 it's kind of a non-salient expense. 00:13:52.630 --> 00:13:55.350 So you don't think about it every time 00:13:55.350 --> 00:13:59.770 you're tapping as much as having value deducted from the card 00:13:59.770 --> 00:14:00.280 every time. 00:14:00.280 --> 00:14:04.732 So those people might ride opportunistically 00:14:04.732 --> 00:14:06.190 because they've already paid for it 00:14:06.190 --> 00:14:09.460 and they might do shorter trips, discretionary trips 00:14:09.460 --> 00:14:11.680 more so than someone who is paying 00:14:11.680 --> 00:14:14.524 every fare independently. 00:14:14.524 --> 00:14:15.940 And another big issue is investing 00:14:15.940 --> 00:14:17.050 in new fare technology. 00:14:17.050 --> 00:14:19.090 That's a headache, a big headache. 00:14:19.090 --> 00:14:23.610 You can imagine changing the way the old technology, all 00:14:23.610 --> 00:14:28.090 the analytics and administration depends on uncertain system. 00:14:28.090 --> 00:14:31.840 And bringing in a new system that nobody understands yet 00:14:31.840 --> 00:14:33.410 might work or not. 00:14:33.410 --> 00:14:36.220 So all the maintenance needs to be figured out. 00:14:36.220 --> 00:14:39.310 And you have to communicate with the public about what impact 00:14:39.310 --> 00:14:42.760 that has on them, and that could involve changes 00:14:42.760 --> 00:14:45.760 in the fare structure, not just the technology. 00:14:45.760 --> 00:14:48.820 So sometimes technology enables new fare structures or new fare 00:14:48.820 --> 00:14:49.890 policies. 00:14:49.890 --> 00:14:51.400 And that can be one of the reasons 00:14:51.400 --> 00:14:54.130 for bringing new technology. 00:14:54.130 --> 00:14:55.844 Questions on the issues? 00:14:55.844 --> 00:14:56.475 [INAUDIBLE] 00:14:56.475 --> 00:14:57.100 AUDIENCE: Yeah. 00:14:57.100 --> 00:15:02.370 I wonder if the rising fare issue 00:15:02.370 --> 00:15:06.620 is less common in privately operated system 00:15:06.620 --> 00:15:07.954 than in public operated systems. 00:15:07.954 --> 00:15:10.161 GABRIEL SANCHEZ-MARTINEZ: Can you give me an example? 00:15:10.161 --> 00:15:11.910 I'm not sure I follow the question. 00:15:11.910 --> 00:15:14.680 AUDIENCE: In systems where bus and rail are exceptions 00:15:14.680 --> 00:15:15.717 to the private sector. 00:15:15.717 --> 00:15:17.050 GABRIEL SANCHEZ-MARTINEZ: Right. 00:15:17.050 --> 00:15:19.390 Sometimes the public sector is the one-- 00:15:19.390 --> 00:15:22.960 so if you look at the case of London, 00:15:22.960 --> 00:15:25.930 for example, even though buses are operated 00:15:25.930 --> 00:15:31.060 by private companies in some way it's a contract 00:15:31.060 --> 00:15:34.150 so the fare policy is still set by TFL, 00:15:34.150 --> 00:15:36.430 which is the government. 00:15:36.430 --> 00:15:38.470 That's I think usually the case if you 00:15:38.470 --> 00:15:45.520 have a push for a unified single brand of the system. 00:15:45.520 --> 00:15:46.930 Maybe in other countries-- 00:15:46.930 --> 00:15:50.110 I'm thinking more of Latin America, Africa-- 00:15:50.110 --> 00:15:51.790 there are sort of independent people 00:15:51.790 --> 00:15:53.750 running their own systems. 00:15:53.750 --> 00:15:56.200 And those are usually cash-based so there's 00:15:56.200 --> 00:15:58.740 no integration whatsoever. 00:15:58.740 --> 00:16:00.585 Maybe you have a different example. 00:16:00.585 --> 00:16:01.380 AUDIENCE: But No. 00:16:01.380 --> 00:16:05.750 Still [INAUDIBLE] just [INAUDIBLE] the government. 00:16:05.750 --> 00:16:09.440 So I wonder if the type of contracts [INAUDIBLE] 00:16:09.440 --> 00:16:12.590 engage with the private [INAUDIBLE] concessions. 00:16:12.590 --> 00:16:13.900 I mean, there is a difference. 00:16:13.900 --> 00:16:16.640 There is a myriad of types of [INAUDIBLE].. 00:16:16.640 --> 00:16:22.820 I wonder if that allows you more flexibility to sort of delay 00:16:22.820 --> 00:16:25.580 because you're under a contract with a specific [INAUDIBLE] 00:16:25.580 --> 00:16:28.420 negotiation [INAUDIBLE]. 00:16:28.420 --> 00:16:32.480 So I wonder if that give you leverage to delay increasing 00:16:32.480 --> 00:16:35.135 fares, as opposed to-- 00:16:35.135 --> 00:16:36.760 GABRIEL SANCHEZ-MARTINEZ: I'm not sure. 00:16:36.760 --> 00:16:38.260 Frankly, I'm not sure. 00:16:38.260 --> 00:16:43.210 Sort of the issues related to concessions in that case 00:16:43.210 --> 00:16:45.700 depend a lot on the contract structure. 00:16:45.700 --> 00:16:52.300 So some concessions derive revenue from the fare. 00:16:52.300 --> 00:16:55.400 I think that's not the best way of doing it. 00:16:55.400 --> 00:16:57.370 But if that's how the contract is structured, 00:16:57.370 --> 00:16:59.930 essentially, you say, I want you to operate. 00:16:59.930 --> 00:17:01.870 You keep the fares. 00:17:01.870 --> 00:17:06.430 Then of course, any fare question will affect revenues 00:17:06.430 --> 00:17:09.250 and it will affect ridership too. 00:17:09.250 --> 00:17:13.510 So the private operator will be interested in any kind of fare 00:17:13.510 --> 00:17:16.900 policy change if that increases their revenues 00:17:16.900 --> 00:17:22.210 and will be disinterested in any fare policy change that 00:17:22.210 --> 00:17:24.760 decreases their revenues, or increases the risk of revenue 00:17:24.760 --> 00:17:25.260 anyway. 00:17:25.260 --> 00:17:28.750 So that is probably a poisonous topic 00:17:28.750 --> 00:17:30.640 during the length of a contract and you 00:17:30.640 --> 00:17:34.780 might have to then do it between contracts. 00:17:34.780 --> 00:17:37.090 Here I'm speculating, but I imagine 00:17:37.090 --> 00:17:42.800 this is a case where you pay to run by distance 00:17:42.800 --> 00:17:45.070 or you might have some performance standards as well, 00:17:45.070 --> 00:17:46.660 but it's not fare so much. 00:17:46.660 --> 00:17:48.580 So the government keeps the fares 00:17:48.580 --> 00:17:51.490 and then pays the operators the contracted amount. 00:17:51.490 --> 00:17:56.770 Then that disconnects the operator's performance 00:17:56.770 --> 00:18:00.820 from fare revenue and it allows the government or the planning 00:18:00.820 --> 00:18:06.140 agency to change the fare structure, change fare 00:18:06.140 --> 00:18:08.960 technology in the middle of a contract. 00:18:08.960 --> 00:18:12.370 And so in the former case, I imagine it's quite complicated. 00:18:12.370 --> 00:18:19.420 Because usually, the contracts are not awarded on cycle, i.e., 00:18:19.420 --> 00:18:22.000 at a certain moment in time you might have a contract that 00:18:22.000 --> 00:18:25.100 expires next month; another one that expires in a year; 00:18:25.100 --> 00:18:27.010 another one that expires in five years. 00:18:27.010 --> 00:18:29.710 So to implement a system-wide change, 00:18:29.710 --> 00:18:32.010 it would be very difficult if the contracts 00:18:32.010 --> 00:18:34.945 depend on the fare revenues. 00:18:38.000 --> 00:18:38.536 Yeah. 00:18:38.536 --> 00:18:39.590 AUDIENCE: So a couple of questions. 00:18:39.590 --> 00:18:42.000 One, it seems like one of the biggest choices you make 00:18:42.000 --> 00:18:44.550 is whether to have [INAUDIBLE] whether to have 00:18:44.550 --> 00:18:45.710 a flat or a zoned fare. 00:18:45.710 --> 00:18:45.940 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:18:45.940 --> 00:18:48.023 And we're going to talk about market segmentation. 00:18:48.023 --> 00:18:48.780 AUDIENCE: OK. 00:18:48.780 --> 00:18:55.810 And then about raising fares, I see that the MTA in New York 00:18:55.810 --> 00:18:58.300 battles to raise fares, so I think 00:18:58.300 --> 00:19:00.950 flow is really hard for them, whereas when I lived London, 00:19:00.950 --> 00:19:03.230 literally every single year on January 1st, the fare changed. 00:19:03.230 --> 00:19:03.390 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:19:03.390 --> 00:19:05.370 So that goes back this sort of annually 00:19:05.370 --> 00:19:08.130 with inflation and strategy where you do it every year 00:19:08.130 --> 00:19:10.877 and it just happens on January 1st. 00:19:10.877 --> 00:19:12.960 AUDIENCE: People get a big grumpy, but [INAUDIBLE] 00:19:12.960 --> 00:19:13.500 about it. 00:19:13.500 --> 00:19:14.833 GABRIEL SANCHEZ-MARTINEZ: Right. 00:19:14.833 --> 00:19:16.242 Right. 00:19:16.242 --> 00:19:17.760 And there are many other examples 00:19:17.760 --> 00:19:20.490 of that kind of approach. 00:19:20.490 --> 00:19:22.250 I think it's more effective, frankly. 00:19:22.250 --> 00:19:26.520 Expenses go up every year, so why not? 00:19:26.520 --> 00:19:30.180 Communicate what the strategy is, make it fair-- 00:19:30.180 --> 00:19:37.710 no pun intended-- and do it. 00:19:37.710 --> 00:19:41.820 AUDIENCE: I read once that King County Transit in Seattle 00:19:41.820 --> 00:19:44.930 discounts fares for people living 00:19:44.930 --> 00:19:50.996 below the poverty line or some threshold based on that. 00:19:50.996 --> 00:19:54.100 [INAUDIBLE] Is that pretty common? 00:19:54.100 --> 00:19:55.350 GABRIEL SANCHEZ-MARTINEZ: Yes. 00:19:55.350 --> 00:19:57.067 So we'll talk-- in market segmentation, 00:19:57.067 --> 00:19:58.650 one of the things you might segment by 00:19:58.650 --> 00:20:01.020 is socioeconomic status. 00:20:01.020 --> 00:20:02.270 Yeah. 00:20:02.270 --> 00:20:04.000 And there's different ways of doing that. 00:20:04.000 --> 00:20:06.120 So we can talk more about that when we get there. 00:20:06.120 --> 00:20:07.842 AUDIENCE: I have another question. 00:20:07.842 --> 00:20:11.830 I also have heard it argued that with the switch 00:20:11.830 --> 00:20:13.890 to all electronic tolling that it 00:20:13.890 --> 00:20:17.879 increases [INAUDIBLE] obscures how much you actually pay. 00:20:17.879 --> 00:20:19.170 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:20:19.170 --> 00:20:21.070 This is a saliency issue. 00:20:21.070 --> 00:20:24.149 AUDIENCE: [INAUDIBLE] switch from cash-base fare system to-- 00:20:24.149 --> 00:20:25.690 GABRIEL SANCHEZ-MARTINEZ: Absolutely. 00:20:25.690 --> 00:20:26.190 Yeah. 00:20:26.190 --> 00:20:28.860 Especially if you go to passes or a capping scheme, 00:20:28.860 --> 00:20:30.450 which we'll discuss. 00:20:30.450 --> 00:20:31.230 So yeah. 00:20:31.230 --> 00:20:33.240 The more obscure, you make the fare system, 00:20:33.240 --> 00:20:39.560 the less of an impact it will have on demand response. 00:20:39.560 --> 00:20:43.630 And that is true in other areas. 00:20:43.630 --> 00:20:46.080 If you look at it from a macroeconomic perspective, 00:20:46.080 --> 00:20:47.080 that's true in business. 00:20:47.080 --> 00:20:48.935 AUDIENCE: Wouldn't that mean the more obscure you 00:20:48.935 --> 00:20:49.870 make it, the less-- 00:20:49.870 --> 00:20:52.286 GABRIEL SANCHEZ-MARTINEZ: Well, if people don't understand 00:20:52.286 --> 00:20:59.350 pricing, it may be you might get a sort of lower reaction 00:20:59.350 --> 00:21:02.410 to changes in pricing. 00:21:02.410 --> 00:21:04.070 You make less of a fuss about it. 00:21:08.710 --> 00:21:09.210 OK. 00:21:09.210 --> 00:21:11.690 Fare structure, so market segmentation. 00:21:11.690 --> 00:21:13.580 That's a good segue with these questions. 00:21:13.580 --> 00:21:17.580 So one possible way of having a fare structure 00:21:17.580 --> 00:21:19.230 is to have a flat fare structure. 00:21:19.230 --> 00:21:20.690 That's the simplest one. 00:21:20.690 --> 00:21:23.970 It costs some amount to ride, no matter what. 00:21:23.970 --> 00:21:26.390 No matter where from, what time of day, 00:21:26.390 --> 00:21:29.190 what mode, that's the price. 00:21:29.190 --> 00:21:30.950 So again, that's the simplest. 00:21:30.950 --> 00:21:33.230 I think most agencies, if you look 00:21:33.230 --> 00:21:35.480 at the profile of all agencies, including small ones, 00:21:35.480 --> 00:21:36.980 have that fare structure. 00:21:36.980 --> 00:21:40.070 Bigger agencies tend to have some market segmentation 00:21:40.070 --> 00:21:46.460 because they are bigger and they serve a wider range of people, 00:21:46.460 --> 00:21:48.530 they serve a wider geographic area, 00:21:48.530 --> 00:21:50.670 and they have some congestion problems. 00:21:50.670 --> 00:21:52.880 So how can we differentiate fare, 00:21:52.880 --> 00:21:54.620 how can we segment the market? 00:21:54.620 --> 00:21:56.160 One is spatially. 00:21:56.160 --> 00:22:00.650 So you could have a flat geographic structure 00:22:00.650 --> 00:22:04.250 where no matter where you ride from, it costs the same. 00:22:04.250 --> 00:22:05.510 You could have zones. 00:22:05.510 --> 00:22:11.870 So here in Boston, we have no zones outside of commuter rail. 00:22:11.870 --> 00:22:13.200 We have no zones. 00:22:13.200 --> 00:22:17.660 When you go to London, there's zones one through nine, 00:22:17.660 --> 00:22:21.630 but most of the greater London is in the first three, about. 00:22:21.630 --> 00:22:28.430 So it costs more to cross zones than to ride within a zone. 00:22:28.430 --> 00:22:34.110 And you can buy passes that are zone 1, zone 1-2, et cetera. 00:22:34.110 --> 00:22:36.770 So that's one way of doing it. 00:22:36.770 --> 00:22:40.880 If you take that to the extreme and imagine many, many zones, 00:22:40.880 --> 00:22:43.790 you could end up with something called distance-based pricing, 00:22:43.790 --> 00:22:46.400 where each of the pair has a different fare, 00:22:46.400 --> 00:22:51.580 depending on how long that ride was. 00:22:51.580 --> 00:22:55.090 So you can do that if you have a tap-in, tap-out system, 00:22:55.090 --> 00:22:58.060 you can do that with proof of payment in some cases. 00:22:58.060 --> 00:23:01.480 It's more common in Europe. 00:23:01.480 --> 00:23:04.270 And regarding zones, a lot of agencies 00:23:04.270 --> 00:23:06.520 used to have more zones and are going 00:23:06.520 --> 00:23:09.370 for reducing zones and simplifying the system. 00:23:09.370 --> 00:23:11.890 It's an interesting, curious trend 00:23:11.890 --> 00:23:14.920 because they've been doing this at the same time 00:23:14.920 --> 00:23:17.350 that they've been investing in new fare technology that 00:23:17.350 --> 00:23:20.920 enables more complex fare structures. 00:23:20.920 --> 00:23:23.920 But it's a choice of communicating to the public 00:23:23.920 --> 00:23:26.920 and keeping it simple, easier to understand. 00:23:26.920 --> 00:23:29.350 AUDIENCE: And do a couple of years' long advertising 00:23:29.350 --> 00:23:30.850 campaign to get people to figure out 00:23:30.850 --> 00:23:33.183 how to cap for the overground system, which 00:23:33.183 --> 00:23:35.020 is part of the TFL. 00:23:35.020 --> 00:23:38.310 Because you could essentially go around the entire city just 00:23:38.310 --> 00:23:40.090 in zone 2. 00:23:40.090 --> 00:23:41.757 But if you don't tap in in the right way 00:23:41.757 --> 00:23:43.340 when you transfer trains or something, 00:23:43.340 --> 00:23:45.220 it will think you went into the city and out. 00:23:45.220 --> 00:23:47.710 It's a lot cheaper to go regularly than it is to go-- 00:23:47.710 --> 00:23:48.070 GABRIEL SANCHEZ-MARTINEZ: Right. 00:23:48.070 --> 00:23:49.310 Because you stay in one zone. 00:23:49.310 --> 00:23:50.560 AUDIENCE: It was so confusing. 00:23:50.560 --> 00:23:52.518 They had people on the platforms explaining it. 00:23:56.355 --> 00:23:59.800 AUDIENCE: Is DC Metro an example of distance-based? 00:23:59.800 --> 00:24:01.930 GABRIEL SANCHEZ-MARTINEZ: In some ways, yes. 00:24:01.930 --> 00:24:07.030 They have many sort of things that influence price. 00:24:07.030 --> 00:24:10.300 So I don't know if you've had the experience of going to CRB 00:24:10.300 --> 00:24:12.370 and trying to figure out how much money to put 00:24:12.370 --> 00:24:16.600 on your smart trip card because it varies by time of day 00:24:16.600 --> 00:24:18.477 or how long you are going to travel. 00:24:18.477 --> 00:24:20.560 And you don't necessarily know that ahead of time, 00:24:20.560 --> 00:24:22.180 so you end up loading-- 00:24:22.180 --> 00:24:23.320 it works in their favor. 00:24:23.320 --> 00:24:26.470 I just load up $20 or $40 depending on-- 00:24:26.470 --> 00:24:29.590 I don't really figure out how much my rides will cost and put 00:24:29.590 --> 00:24:31.510 in just that amount. 00:24:31.510 --> 00:24:34.330 Eventually, maybe I lose my card and they keep the revenue. 00:24:36.920 --> 00:24:37.420 OK. 00:24:37.420 --> 00:24:38.630 Another-- oh, Eli. 00:24:38.630 --> 00:24:40.820 AUDIENCE: So you said that you really need 00:24:40.820 --> 00:24:42.653 a tap-in, tap-out system for distance-based, 00:24:42.653 --> 00:24:44.944 but for zonal [INAUDIBLE],, how do you run that? 00:24:44.944 --> 00:24:46.360 GABRIEL SANCHEZ-MARTINEZ: It could 00:24:46.360 --> 00:24:50.640 be proof of payment as well, not just check-in, though. 00:24:50.640 --> 00:24:53.850 So in Europe, you have some symptoms, 00:24:53.850 --> 00:24:56.680 especially it happens more intercity. 00:24:56.680 --> 00:25:02.140 So you buy a ticket between this station and that station 00:25:02.140 --> 00:25:04.650 and you have a paper ticket. 00:25:04.650 --> 00:25:08.110 And you when you board you validate it so essentially, 00:25:08.110 --> 00:25:11.260 it gets canceled with a timestamp. 00:25:11.260 --> 00:25:15.160 And if a fare inspector comes, you better have one. 00:25:15.160 --> 00:25:17.840 Otherwise, you'll be heavily fined. 00:25:17.840 --> 00:25:20.170 So that's one way of doing it. 00:25:23.190 --> 00:25:24.870 Any other questions? 00:25:24.870 --> 00:25:25.510 OK. 00:25:25.510 --> 00:25:28.370 Temporal segmentation, so that's pretty straightforward. 00:25:28.370 --> 00:25:30.160 You've got a flat system where there 00:25:30.160 --> 00:25:33.100 is no change in fares across time of day 00:25:33.100 --> 00:25:35.890 and another one where you have a peak surcharge 00:25:35.890 --> 00:25:37.300 or an off-peak discount. 00:25:37.300 --> 00:25:39.700 Same thing, two ways of marketing it. 00:25:39.700 --> 00:25:43.390 Most agencies prefer to offer a discount on off-peak 00:25:43.390 --> 00:25:46.090 than to charge more during the peak, 00:25:46.090 --> 00:25:48.350 but it's all psychological. 00:25:48.350 --> 00:25:51.640 So why would an agency be interested in doing 00:25:51.640 --> 00:25:55.090 this and what are the impacts? 00:25:55.090 --> 00:25:55.590 Henry. 00:25:55.590 --> 00:25:58.530 AUDIENCE: It would influence people to maybe think more 00:25:58.530 --> 00:26:02.660 about taking the trips during off-peak so that there are 00:26:02.660 --> 00:26:04.300 alternatives that they-- 00:26:04.300 --> 00:26:05.508 GABRIEL SANCHEZ-MARTINEZ: OK. 00:26:05.508 --> 00:26:10.270 So nudging people who can do off-peak to do off-peak 00:26:10.270 --> 00:26:13.300 rather than during the peak. 00:26:13.300 --> 00:26:14.120 What else? 00:26:14.120 --> 00:26:17.040 Any other ideas? 00:26:17.040 --> 00:26:19.250 AUDIENCE: Just encourage more off-peak ridership 00:26:19.250 --> 00:26:21.476 in general, like people should probably do this. 00:26:21.476 --> 00:26:21.718 GABRIEL SANCHEZ-MARTINEZ: Right. 00:26:21.718 --> 00:26:23.310 So more discretionary trips. 00:26:23.310 --> 00:26:26.331 So encourage more off-peak ridership. 00:26:26.331 --> 00:26:26.830 OK. 00:26:26.830 --> 00:26:29.200 And why does it make sense to charge more 00:26:29.200 --> 00:26:31.690 during the peak than off-peak, regardless 00:26:31.690 --> 00:26:33.492 of how you market it? 00:26:33.492 --> 00:26:35.436 AUDIENCE: Some people are [INAUDIBLE] flexible 00:26:35.436 --> 00:26:38.484 in their schedules and [INAUDIBLE] capture that. 00:26:38.484 --> 00:26:39.900 GABRIEL SANCHEZ-MARTINEZ: Exactly. 00:26:39.900 --> 00:26:41.850 So willingness to pay is higher, either 00:26:41.850 --> 00:26:43.230 because they are actually willing 00:26:43.230 --> 00:26:45.600 or because they are captured on time like they 00:26:45.600 --> 00:26:49.350 have to arrive at some time to work so they 00:26:49.350 --> 00:26:50.520 don't have many options. 00:26:50.520 --> 00:26:53.760 And that means that they're much less flexible 00:26:53.760 --> 00:26:58.960 and that their response will be much lower to a price change. 00:26:58.960 --> 00:27:01.340 OK. 00:27:01.340 --> 00:27:03.880 By service, that's pretty simple too. 00:27:03.880 --> 00:27:05.630 You could charge the same for bus and rail 00:27:05.630 --> 00:27:07.850 or you could say rail is a better service so it's 00:27:07.850 --> 00:27:09.120 going to be more expensive. 00:27:09.120 --> 00:27:10.910 The MBTA has this option. 00:27:10.910 --> 00:27:13.970 Regular bus versus express bus, another way in which 00:27:13.970 --> 00:27:16.370 the MBTA segments the market. 00:27:16.370 --> 00:27:20.990 So the idea is if it's a nicer service, you pay more. 00:27:20.990 --> 00:27:22.520 By socioeconomic status. 00:27:22.520 --> 00:27:24.710 So students might get a discount, seniors, 00:27:24.710 --> 00:27:26.900 disabled might get a discount. 00:27:26.900 --> 00:27:29.390 We have some rules here where seniors 00:27:29.390 --> 00:27:32.520 have to be offered half the base fare, 00:27:32.520 --> 00:27:35.590 so that's one example of that. 00:27:35.590 --> 00:27:37.620 Some of it can be mandated by the government. 00:27:37.620 --> 00:27:38.995 And then there's social programs. 00:27:38.995 --> 00:27:41.000 So there's different ways of doing that. 00:27:41.000 --> 00:27:43.070 It could be as part of your participation 00:27:43.070 --> 00:27:45.800 in some social program, it could also 00:27:45.800 --> 00:27:51.560 be a needs-based subsidy where you essentially 00:27:51.560 --> 00:27:56.730 prove that you have financial need and based on your need, 00:27:56.730 --> 00:27:58.370 you give some amount of subsidy. 00:27:58.370 --> 00:28:02.000 So a subsidy could be on or off, you either have the subsidy 00:28:02.000 --> 00:28:05.520 or you don't or it could be depending on your need, 00:28:05.520 --> 00:28:07.670 you have more subsidy or less subsidy. 00:28:07.670 --> 00:28:10.280 And usually, it's the simpler one. 00:28:10.280 --> 00:28:13.580 I don't actually know of any example that 00:28:13.580 --> 00:28:16.732 is done based on your specific case, 00:28:16.732 --> 00:28:18.440 but you could think of doing it that way. 00:28:18.440 --> 00:28:23.890 Because if somebody is very, very poor then 00:28:23.890 --> 00:28:26.910 they might need more subsidy than somebody who 00:28:26.910 --> 00:28:30.124 is just below the poverty line. 00:28:30.124 --> 00:28:34.108 AUDIENCE: [INAUDIBLE] is that because those groups tend 00:28:34.108 --> 00:28:36.754 to have less ability to pay or what's the logic? 00:28:36.754 --> 00:28:38.920 GABRIEL SANCHEZ-MARTINEZ: That is usually the logic. 00:28:38.920 --> 00:28:40.560 So yeah. 00:28:40.560 --> 00:28:44.130 Again, it's sort of social equity goal. 00:28:44.130 --> 00:28:50.380 So seniors and students often can't drive, et cetera. 00:28:50.380 --> 00:28:53.460 So it's the goal of the transit agency or the transit provider 00:28:53.460 --> 00:28:58.185 to increase mobility for all groups of people, 00:28:58.185 --> 00:29:01.210 the way in which that goal is satisfied. 00:29:01.210 --> 00:29:01.945 Henry. 00:29:01.945 --> 00:29:03.805 AUDIENCE: How about programs where 00:29:03.805 --> 00:29:09.060 you can take out your income pre-tax and [INAUDIBLE]?? 00:29:09.060 --> 00:29:11.270 Is that like a social program? 00:29:11.270 --> 00:29:12.850 GABRIEL SANCHEZ-MARTINEZ: No. 00:29:12.850 --> 00:29:16.220 I do have that in another lecture, yeah. 00:29:16.220 --> 00:29:18.260 Because that is done for people-- 00:29:18.260 --> 00:29:22.610 often, they are employee who have a decent salary. 00:29:22.610 --> 00:29:25.000 So that's an incentive. 00:29:25.000 --> 00:29:28.460 And that incentive exists for a car as well. 00:29:28.460 --> 00:29:32.630 So that's a different strategy. 00:29:32.630 --> 00:29:34.910 Another way of segmenting is by willingness to prepay. 00:29:34.910 --> 00:29:39.110 So in other words, are you willing to buy a pass? 00:29:39.110 --> 00:29:41.420 So if you can commit to a whole month, 00:29:41.420 --> 00:29:43.280 then you get some discount. 00:29:43.280 --> 00:29:47.840 And this depends heavily on what the break even point is. 00:29:47.840 --> 00:29:50.830 Usually, passes-- it used to be the case that passes 00:29:50.830 --> 00:29:52.830 had heavy discounts. 00:29:52.830 --> 00:29:59.210 So even without having a commute two times a week for five days, 00:29:59.210 --> 00:30:01.010 you would still get a discount. 00:30:01.010 --> 00:30:05.690 Now it tends to require sometimes between two 00:30:05.690 --> 00:30:09.050 and three trips daily to get that break even point. 00:30:09.050 --> 00:30:11.870 Some people still decide to go with that because it's more 00:30:11.870 --> 00:30:14.570 convenient, it's less salient. 00:30:14.570 --> 00:30:20.300 So it's another way of discriminating or segmenting 00:30:20.300 --> 00:30:22.240 the market. 00:30:22.240 --> 00:30:24.730 Any questions on market segmentation? 00:30:24.730 --> 00:30:26.800 AUDIENCE: [INAUDIBLE] so speaking 00:30:26.800 --> 00:30:31.439 of [INAUDIBLE],, the subsidies given by government? 00:30:31.439 --> 00:30:32.980 GABRIEL SANCHEZ-MARTINEZ: It depends. 00:30:32.980 --> 00:30:35.020 So it could be that the agency has 00:30:35.020 --> 00:30:37.570 a mandate they have to do this. 00:30:37.570 --> 00:30:40.630 That is often the model. 00:30:40.630 --> 00:30:42.520 From an agency's perspective it would 00:30:42.520 --> 00:30:47.680 be much preferable to have the subsidy come from someone else. 00:30:47.680 --> 00:30:49.540 Because then that would allow the agency 00:30:49.540 --> 00:30:52.390 to focus on operating, It would set fares 00:30:52.390 --> 00:30:56.140 to what it needs to cover operational expenses. 00:30:56.140 --> 00:30:59.110 And then if it's a social program, well, 00:30:59.110 --> 00:31:02.200 why doesn't that social program provide the subsidy? 00:31:02.200 --> 00:31:09.370 If it's disabled, maybe there's some other government entity 00:31:09.370 --> 00:31:11.560 that can provide that subsidy. 00:31:11.560 --> 00:31:14.200 For students, why not the schools 00:31:14.200 --> 00:31:16.240 or the Department of Education. 00:31:16.240 --> 00:31:19.420 So that's an interesting sort of policy debate 00:31:19.420 --> 00:31:21.700 and we could talk about it for a long time 00:31:21.700 --> 00:31:23.890 because some people are very much against it 00:31:23.890 --> 00:31:27.300 and some people think that's a good idea. 00:31:27.300 --> 00:31:28.110 Yeah. 00:31:28.110 --> 00:31:31.330 AUDIENCE: Also another thing about the various groups that 00:31:31.330 --> 00:31:33.050 are entitled to subsidies is that you 00:31:33.050 --> 00:31:36.160 have to have some mechanism within your technology 00:31:36.160 --> 00:31:38.179 that verifies that a person is [INAUDIBLE].. 00:31:38.179 --> 00:31:39.970 GABRIEL SANCHEZ-MARTINEZ: To prevent fraud. 00:31:39.970 --> 00:31:41.660 Yeah. 00:31:41.660 --> 00:31:45.230 So smart cards are helpful with that 00:31:45.230 --> 00:31:51.140 because you can embed the transit chip into a student 00:31:51.140 --> 00:31:52.360 card, for example. 00:31:52.360 --> 00:31:54.970 And that's what we have here at MIT. 00:31:54.970 --> 00:31:55.470 OK. 00:31:55.470 --> 00:31:56.400 Fare elasticities. 00:31:56.400 --> 00:31:59.850 So they can range from -1 to 0. 00:31:59.850 --> 00:32:04.320 We know that demand of transit is inelastic. 00:32:04.320 --> 00:32:10.860 So it means that almost always, virtually in every case, 00:32:10.860 --> 00:32:13.020 if you raise fares, you increase your revenue 00:32:13.020 --> 00:32:15.560 and your ridership decreases not by too much, 00:32:15.560 --> 00:32:18.370 not by so much that your fare revenues go down. 00:32:18.370 --> 00:32:22.680 So here are some examples of typical elasticities. 00:32:22.680 --> 00:32:26.460 They usually are closer to -0.3 or -0.4. 00:32:26.460 --> 00:32:32.130 For rail, it's a little less, so -0.2, -0.15. 00:32:32.130 --> 00:32:35.010 For off-peak, it tends to be twice as much 00:32:35.010 --> 00:32:37.830 because more trips are discretionary. 00:32:37.830 --> 00:32:40.440 For work trips, it tends to be much more inelastic, 00:32:40.440 --> 00:32:43.890 as we discussed, so even half of their base one. 00:32:43.890 --> 00:32:45.600 And there's something interesting 00:32:45.600 --> 00:32:47.560 that is maybe not as important anymore, 00:32:47.560 --> 00:32:49.562 but I wanted to add it here. 00:32:49.562 --> 00:32:51.270 There is a higher demand for free transit 00:32:51.270 --> 00:32:53.190 than there is for very, very cheap transit. 00:32:53.190 --> 00:32:54.720 Even if you have to pay, say a cent, 00:32:54.720 --> 00:33:00.570 to really get to the extreme, the simplicity of not having 00:33:00.570 --> 00:33:03.810 to deal with paying fares, not having to interact with a fare 00:33:03.810 --> 00:33:07.650 system, and from the marketing point of view, it's free-- 00:33:07.650 --> 00:33:11.850 you can just take it whenever, you don't 00:33:11.850 --> 00:33:14.790 have to be ready to take it-- 00:33:14.790 --> 00:33:17.790 then that sort of drives demand up. 00:33:17.790 --> 00:33:19.680 So that's an interesting finding. 00:33:19.680 --> 00:33:21.510 A lot of agencies have moved away from that 00:33:21.510 --> 00:33:24.150 because these are usually smaller agencies. 00:33:24.150 --> 00:33:28.319 But of course, operational costs increase and at some point, 00:33:28.319 --> 00:33:29.610 you need to cover them somehow. 00:33:32.210 --> 00:33:39.990 So these inelastic values mean that raising fares 00:33:39.990 --> 00:33:42.082 is an effective way of increasing revenues. 00:33:42.082 --> 00:33:43.540 Not so much to increase the demand, 00:33:43.540 --> 00:33:46.710 although sometimes some aspects of fare policy 00:33:46.710 --> 00:33:48.300 are changed to increase the amount. 00:33:48.300 --> 00:33:51.270 You know you're going to get a small response to that. 00:33:51.270 --> 00:33:54.300 From a microeconomics perspective, 00:33:54.300 --> 00:33:56.880 if you think about market segmentation and running 00:33:56.880 --> 00:33:58.920 transit like a business, then you 00:33:58.920 --> 00:34:02.760 would want to charge more for longer trips; 00:34:02.760 --> 00:34:05.640 trips in more convenient, reliable, comfortable, 00:34:05.640 --> 00:34:09.420 and frequent modes; peak period trips; trips where 00:34:09.420 --> 00:34:12.750 other modes are inconvenient or costly, therefore this 00:34:12.750 --> 00:34:15.030 is the best option and why not charge more for it; 00:34:15.030 --> 00:34:17.250 and trips subsidized by third parties 00:34:17.250 --> 00:34:20.639 because then you could capture some of the increased 00:34:20.639 --> 00:34:22.590 fare from an outside source. 00:34:22.590 --> 00:34:23.159 All right. 00:34:23.159 --> 00:34:27.060 So of course, not everybody would 00:34:27.060 --> 00:34:29.730 agree that this is what agencies should do. 00:34:29.730 --> 00:34:31.230 But if you are thinking of running 00:34:31.230 --> 00:34:34.170 an agency as a business, this is exactly what you want. 00:34:34.170 --> 00:34:37.530 If you have more willingness to pay for whatever reason, 00:34:37.530 --> 00:34:39.139 your fare is higher. 00:34:39.139 --> 00:34:41.760 And that means that fare revenues go up 00:34:41.760 --> 00:34:43.969 and your fare recovery ratio goes up. 00:34:43.969 --> 00:34:46.949 And that means you can operate more service 00:34:46.949 --> 00:34:48.780 or you can operate more reliably or you can 00:34:48.780 --> 00:34:50.591 get nicer vehicles, et cetera. 00:34:50.591 --> 00:34:51.090 All right. 00:34:53.880 --> 00:34:55.900 So transfer pricing and policy. 00:34:55.900 --> 00:34:58.320 Another issue, some agencies don't 00:34:58.320 --> 00:35:02.670 have any sort of integration between rides 00:35:02.670 --> 00:35:05.790 so every time you board a vehicle, you pay a full fare 00:35:05.790 --> 00:35:07.710 and there's no free transfer. 00:35:07.710 --> 00:35:10.470 Some agencies have a reduced price transfer 00:35:10.470 --> 00:35:11.880 or a free transfer. 00:35:11.880 --> 00:35:14.950 And free transfers can be done in different ways. 00:35:14.950 --> 00:35:17.250 One of the most common ways is to have 00:35:17.250 --> 00:35:19.590 a time-based free transfer. 00:35:19.590 --> 00:35:23.850 So you tap and you get, say, an hour 00:35:23.850 --> 00:35:29.920 or two hours of free transit starting from that point. 00:35:29.920 --> 00:35:33.270 So if you transfer, you'll presumably 00:35:33.270 --> 00:35:36.450 be in this two-hour window and the next time you tap, 00:35:36.450 --> 00:35:38.490 it'll be a free transfer. 00:35:38.490 --> 00:35:42.390 Here in the MBTA, the price is higher for rail and bus. 00:35:42.390 --> 00:35:45.390 So if you transfer from bus to rail, you pay the difference. 00:35:45.390 --> 00:35:49.440 If you transfer from rail to bus, both are free. 00:35:49.440 --> 00:35:52.110 Interestingly, with this time-based rule, 00:35:52.110 --> 00:35:55.360 you can often do something very quickly and return 00:35:55.360 --> 00:35:57.510 and these are actually two different trips 00:35:57.510 --> 00:36:02.000 and you would still get a free return trip 00:36:02.000 --> 00:36:03.640 and it's not a transfer. 00:36:03.640 --> 00:36:05.460 So with more complex fare rules, you 00:36:05.460 --> 00:36:09.270 could say you should get free transfers within this window, 00:36:09.270 --> 00:36:11.340 but not if you are on the same bus route 00:36:11.340 --> 00:36:14.220 or if you are tapping into the same station or things 00:36:14.220 --> 00:36:15.150 like this. 00:36:15.150 --> 00:36:17.880 You could think of the rules used 00:36:17.880 --> 00:36:21.880 for transfer insurance when we discussed all the [INAUDIBLE].. 00:36:21.880 --> 00:36:23.940 These are not fare transfers. 00:36:23.940 --> 00:36:26.260 Those were sort of a logical trip transfers 00:36:26.260 --> 00:36:27.900 and you could apply those rules. 00:36:27.900 --> 00:36:29.640 You could put them into a fare engine 00:36:29.640 --> 00:36:35.160 and not give free transfers when it's not a transfer. 00:36:35.160 --> 00:36:35.670 OK. 00:36:35.670 --> 00:36:37.950 Pay-as-you-go, passes, and capping. 00:36:37.950 --> 00:36:38.450 OK. 00:36:38.450 --> 00:36:40.340 One question before we-- 00:36:40.340 --> 00:36:41.100 Daniel. 00:36:41.100 --> 00:36:43.250 Yeah. 00:36:43.250 --> 00:36:44.770 AUDIENCE: Is there-- how do agencies 00:36:44.770 --> 00:36:48.796 determine if it's unlimited free transfers within the same trip 00:36:48.796 --> 00:36:52.030 or [INAUDIBLE]? 00:36:52.030 --> 00:36:57.088 Because [INAUDIBLE] bus, rail, bus [INAUDIBLE] 00:36:57.088 --> 00:36:58.550 have to [INAUDIBLE]. 00:36:58.550 --> 00:36:59.000 GABRIEL SANCHEZ-MARTINEZ: You usually 00:36:59.000 --> 00:37:01.370 don't have to because they set the window wide enough 00:37:01.370 --> 00:37:05.600 so that you can get your whole three-four trip done 00:37:05.600 --> 00:37:07.121 in that two-hour window. 00:37:07.121 --> 00:37:09.942 AUDIENCE: [INAUDIBLE] trips and I've been charged [INAUDIBLE].. 00:37:09.942 --> 00:37:11.150 GABRIEL SANCHEZ-MARTINEZ: Oh. 00:37:11.150 --> 00:37:14.270 In theory, I think you could go to customer service and say, 00:37:14.270 --> 00:37:15.410 this was a transfer. 00:37:15.410 --> 00:37:17.420 Your policy is free transfers. 00:37:17.420 --> 00:37:21.880 You need to refund me. 00:37:21.880 --> 00:37:22.380 Yeah. 00:37:26.520 --> 00:37:28.230 I don't know the details of it, but I 00:37:28.230 --> 00:37:30.105 imagine that's the case, that you could argue 00:37:30.105 --> 00:37:31.290 that you deserve a refund. 00:37:31.290 --> 00:37:33.266 AUDIENCE: [INAUDIBLE] paid fare policy [INAUDIBLE] 00:37:33.266 --> 00:37:34.200 reach your [INAUDIBLE]. 00:37:34.200 --> 00:37:35.408 GABRIEL SANCHEZ-MARTINEZ: OK. 00:37:35.408 --> 00:37:37.260 I haven't, clearly. 00:37:37.260 --> 00:37:39.630 It would probably be very interesting, actually. 00:37:39.630 --> 00:37:42.196 Any other questions on transfers? 00:37:42.196 --> 00:37:44.070 Transfers are more of an issue when there are 00:37:44.070 --> 00:37:46.137 multiple operators in a region. 00:37:46.137 --> 00:37:47.970 And we'll talk a little bit more about that. 00:37:47.970 --> 00:37:53.890 If it's one operator, that's sort of less controversial. 00:37:53.890 --> 00:37:54.510 OK. 00:37:54.510 --> 00:37:56.850 So pay-as-you-go passes and capping. 00:37:56.850 --> 00:38:02.580 So pay-as-you-go is the simplest way of paying for trips. 00:38:02.580 --> 00:38:05.190 You can do that with cash or with tickets or smart cards 00:38:05.190 --> 00:38:06.670 that have a balance. 00:38:06.670 --> 00:38:09.210 So you load up some value onto the card 00:38:09.210 --> 00:38:14.080 and every time you interact with a fare box or fare gate, 00:38:14.080 --> 00:38:18.480 some amount of that balance is deducted from the card. 00:38:18.480 --> 00:38:20.407 That's pretty clear, hopefully. 00:38:20.407 --> 00:38:21.240 Then there's passes. 00:38:21.240 --> 00:38:23.850 So the idea of a pass is that you sort of prepay 00:38:23.850 --> 00:38:28.650 for three days, for a week, sometimes for a day, 00:38:28.650 --> 00:38:30.060 sometimes for a month. 00:38:30.060 --> 00:38:35.610 And it's a fixed price so you ride with zero marginal costs 00:38:35.610 --> 00:38:37.330 after paying that fixed price. 00:38:37.330 --> 00:38:40.230 If you ride frequently enough, then you break even 00:38:40.230 --> 00:38:42.790 and anything beyond that is a free ride. 00:38:42.790 --> 00:38:48.360 So it's interesting to note that because passes are convenient, 00:38:48.360 --> 00:38:51.630 some fare revenue is derived from past sales 00:38:51.630 --> 00:38:54.210 for people that don't break even. 00:38:54.210 --> 00:38:55.840 That could happen for several reasons. 00:38:55.840 --> 00:38:57.780 One of them might be it's more convenient 00:38:57.780 --> 00:39:00.690 and I'm not going to figure out every month if I'm 00:39:00.690 --> 00:39:03.420 going to be having the pass or not the pass. 00:39:03.420 --> 00:39:06.510 It could also be that some people receive a subsidy 00:39:06.510 --> 00:39:09.660 from somewhere and what they spend 00:39:09.660 --> 00:39:12.126 on the pass is lower than the pass value. 00:39:12.126 --> 00:39:14.250 And so for them, the real break even point is lower 00:39:14.250 --> 00:39:17.580 and you may not necessarily be considering that when 00:39:17.580 --> 00:39:19.681 you do this break even point calculation. 00:39:22.390 --> 00:39:27.072 So in some cases, going back to transfers, agencies 00:39:27.072 --> 00:39:28.530 that don't have free transfers, you 00:39:28.530 --> 00:39:30.870 can get them indirectly through a pass. 00:39:30.870 --> 00:39:33.660 Because they might say, well, you 00:39:33.660 --> 00:39:36.480 break even at 10 rides per week so that's 00:39:36.480 --> 00:39:41.400 your five weekdays times two of going and returning. 00:39:41.400 --> 00:39:44.880 And if you have no free transfers policy 00:39:44.880 --> 00:39:47.610 but your trip requires one transfer, 00:39:47.610 --> 00:39:50.430 then you're going to very quickly reach that limit 00:39:50.430 --> 00:39:52.420 and you're going to ride free after that. 00:39:52.420 --> 00:39:56.590 So in practice, that would be a free transfer policy, 00:39:56.590 --> 00:39:58.770 although not directly. 00:39:58.770 --> 00:40:01.430 So passes increase convenience, we know that. 00:40:01.430 --> 00:40:03.290 The reduce saliency so you're not 00:40:03.290 --> 00:40:05.720 thinking about how much money you're paying every time 00:40:05.720 --> 00:40:07.820 because it's more convenient. 00:40:07.820 --> 00:40:11.240 They can be subsidized by employers, by universities, 00:40:11.240 --> 00:40:11.990 by governments. 00:40:11.990 --> 00:40:14.610 So here we talk about the pre-tax benefit 00:40:14.610 --> 00:40:16.310 that Henry brought up. 00:40:16.310 --> 00:40:18.440 And then there are social programs, 00:40:18.440 --> 00:40:21.890 like access to jobs programs, that give 00:40:21.890 --> 00:40:24.860 passes subsidized or sometimes even free 00:40:24.860 --> 00:40:29.660 for people that are economically disadvantaged. 00:40:29.660 --> 00:40:31.340 Questions on passes? 00:40:31.340 --> 00:40:33.091 I think we're all familiar with passes. 00:40:33.091 --> 00:40:33.590 OK. 00:40:33.590 --> 00:40:36.950 So capping is less common, it's a newer thing. 00:40:36.950 --> 00:40:38.300 It's more common in Europe. 00:40:38.300 --> 00:40:41.400 London has it, Germany has it in many places, the Netherlands. 00:40:41.400 --> 00:40:47.000 So the idea is that when you have a complex fare 00:40:47.000 --> 00:40:51.980 structure where the prices are difficult to figure out, 00:40:51.980 --> 00:40:56.030 then the agency says, don't worry about calculating which 00:40:56.030 --> 00:40:58.910 fare products you need to buy; we'll do the math 00:40:58.910 --> 00:41:01.140 and give you the best price. 00:41:01.140 --> 00:41:03.050 So that's called best price guarantee. 00:41:03.050 --> 00:41:05.480 And the idea is that instead of committing 00:41:05.480 --> 00:41:08.390 to getting a pass or not a pass, you 00:41:08.390 --> 00:41:12.350 pay as you go but when you reach a certain daily limit, 00:41:12.350 --> 00:41:14.550 no more is charged during that day. 00:41:14.550 --> 00:41:15.950 So you can keep riding. 00:41:15.950 --> 00:41:17.670 And that doesn't stop there. 00:41:17.670 --> 00:41:20.450 So the card keeps counting the trips 00:41:20.450 --> 00:41:23.060 in one week, the trips in one month. 00:41:23.060 --> 00:41:25.180 And so you might have multiple caps 00:41:25.180 --> 00:41:29.080 and you might reach different caps on one session 00:41:29.080 --> 00:41:31.280 sort of with one card. 00:41:31.280 --> 00:41:34.220 So essentially, you end up getting 00:41:34.220 --> 00:41:38.570 the equivalent of a pass if it was worth it and not otherwise. 00:41:38.570 --> 00:41:41.300 Obviously, that means agencies lose some revenue 00:41:41.300 --> 00:41:44.640 from people who didn't make the right decision, 00:41:44.640 --> 00:41:50.030 but it's one way of having a complex fare structure that 00:41:50.030 --> 00:41:52.220 heavily segments the market and therefore does 00:41:52.220 --> 00:41:53.870 obtain more fare revenues. 00:41:53.870 --> 00:41:59.930 And it allows for simple communication with people 00:41:59.930 --> 00:42:02.720 by saying, don't worry about it; we know it's complicated, 00:42:02.720 --> 00:42:04.640 but we'll figure it out for you. 00:42:04.640 --> 00:42:08.250 That's one argument. 00:42:08.250 --> 00:42:11.010 In London, one of the reasons capping was introduced 00:42:11.010 --> 00:42:15.030 was with the movement to contactless bank cards. 00:42:15.030 --> 00:42:18.350 So essentially, you can pay with your Visa or MasterCard. 00:42:18.350 --> 00:42:21.570 You can have it on your phone, Apple Pay or Android Pay, 00:42:21.570 --> 00:42:24.550 you tap, it works. 00:42:24.550 --> 00:42:28.830 So obviously, a normal Visa might not 00:42:28.830 --> 00:42:32.070 have the logic inside the card's computer 00:42:32.070 --> 00:42:35.410 to hold a pass or anything like that 00:42:35.410 --> 00:42:37.260 so the card becomes a token. 00:42:37.260 --> 00:42:41.610 One way of dealing with that is to say, it's capped. 00:42:41.610 --> 00:42:43.740 We'll keep track of your usage for the month 00:42:43.740 --> 00:42:45.660 and when you reach the monthly cap, 00:42:45.660 --> 00:42:47.820 we'll stop charging you for it. 00:42:47.820 --> 00:42:52.650 So it simplifies the move to or towards contactless bank card 00:42:52.650 --> 00:42:53.770 usage. 00:42:53.770 --> 00:42:55.210 And we'll talk more about that. 00:42:55.210 --> 00:42:55.490 Yeah. 00:42:55.490 --> 00:42:57.239 AUDIENCE: So since London introduced this, 00:42:57.239 --> 00:43:00.040 have they seen a drop in the number of pass purchases? 00:43:00.040 --> 00:43:01.290 GABRIEL SANCHEZ-MARTINEZ: Yes. 00:43:01.290 --> 00:43:03.270 Well, they've seen an increase in-- 00:43:03.270 --> 00:43:06.270 the penetration for contactless bank card use 00:43:06.270 --> 00:43:07.830 has been increasing steadily. 00:43:07.830 --> 00:43:12.030 And if you go to London, my last experience there 00:43:12.030 --> 00:43:16.500 was on the train ride between the airport 00:43:16.500 --> 00:43:20.670 and the center of London, I heard several people talking 00:43:20.670 --> 00:43:22.050 about how convenient it was. 00:43:22.050 --> 00:43:24.960 So people are excited about this. 00:43:24.960 --> 00:43:27.150 And it's great for a city like London 00:43:27.150 --> 00:43:28.640 that has a lot of people that are 00:43:28.640 --> 00:43:31.950 occasional users, a lot of business travelers being 00:43:31.950 --> 00:43:34.290 able to not have to think about it 00:43:34.290 --> 00:43:36.350 and just pay with a credit card. 00:43:36.350 --> 00:43:38.760 It really helps. 00:43:38.760 --> 00:43:40.230 And it has several other benefits. 00:43:40.230 --> 00:43:43.129 I'll discuss them more in detail when we get to smart cards. 00:43:43.129 --> 00:43:44.670 AUDIENCE: If you say that it's capped 00:43:44.670 --> 00:43:48.080 and you said that there's a trend away from passes, 00:43:48.080 --> 00:43:49.650 people are not buying passes. 00:43:49.650 --> 00:43:51.650 But isn't that essentially saying that actually, 00:43:51.650 --> 00:43:54.120 no; people are buying more passes than ever. 00:43:54.120 --> 00:43:56.900 GABRIEL SANCHEZ-MARTINEZ: Well, it's capping so it's not-- 00:43:56.900 --> 00:43:58.524 AUDIENCE: It's capping, but it comes-- 00:43:58.524 --> 00:44:00.690 GABRIEL SANCHEZ-MARTINEZ: It's equivalent to a pass. 00:44:00.690 --> 00:44:02.340 The math works out so that it a pass, 00:44:02.340 --> 00:44:04.970 but you're not selling a monthly pass. 00:44:04.970 --> 00:44:05.760 Yeah. 00:44:05.760 --> 00:44:07.650 It's not a pass. 00:44:07.650 --> 00:44:08.767 You're paying each ride. 00:44:08.767 --> 00:44:09.840 AUDIENCE: --acts like a [INAUDIBLE].. 00:44:09.840 --> 00:44:10.640 It induces [INAUDIBLE]. 00:44:10.640 --> 00:44:11.264 AUDIENCE: Yeah. 00:44:11.264 --> 00:44:14.340 But I think if you go to London every couple of months 00:44:14.340 --> 00:44:16.330 and you can just show up in London 00:44:16.330 --> 00:44:18.704 and you don't have to fill up an Oyster card or anything. 00:44:18.704 --> 00:44:21.525 You can tap your card and use it like you normally would. 00:44:21.525 --> 00:44:24.060 If you ride too much in a day, they cap it as well, 00:44:24.060 --> 00:44:26.430 which is something they don't do on a monthly pass. 00:44:26.430 --> 00:44:27.721 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:44:29.430 --> 00:44:29.970 OK. 00:44:29.970 --> 00:44:32.280 We'll discuss more on the contactless bank cards. 00:44:32.280 --> 00:44:35.040 This is a very interesting trend that we're seeing. 00:44:35.040 --> 00:44:37.650 So fare policy demand analysis, we talked a little bit 00:44:37.650 --> 00:44:40.080 about this on ridership forecasting. 00:44:40.080 --> 00:44:42.930 Just a little recap, the traditional four-step modeling 00:44:42.930 --> 00:44:44.580 process is not usually appropriate. 00:44:44.580 --> 00:44:48.270 And that's because A, it has insufficient spatiotemporal 00:44:48.270 --> 00:44:52.140 resolution and B, the planning horizon for a fare hike 00:44:52.140 --> 00:44:55.500 is kind of short- to medium-term so you're not 00:44:55.500 --> 00:45:00.120 thinking about land use changes that will generate new trips. 00:45:00.120 --> 00:45:03.300 So therefore, total amount is, we can assume, fixed. 00:45:03.300 --> 00:45:05.970 And what you want to play with is the modal split 00:45:05.970 --> 00:45:10.680 and the assignment to specific trips. 00:45:10.680 --> 00:45:12.930 The other thing is that these four-step models usually 00:45:12.930 --> 00:45:15.810 don't have all the details of all the fare products in them. 00:45:15.810 --> 00:45:17.850 And for an agency, it's important to know 00:45:17.850 --> 00:45:20.010 what people will do about getting a pass 00:45:20.010 --> 00:45:21.600 or paying as they go. 00:45:21.600 --> 00:45:26.040 So usually, fare elasticity analysis is simple; 00:45:26.040 --> 00:45:29.520 in reality, a passenger is considering multiple things 00:45:29.520 --> 00:45:30.400 at the same time. 00:45:30.400 --> 00:45:33.256 So which mode are they going to take, 00:45:33.256 --> 00:45:35.130 are they going to drive their car or are they 00:45:35.130 --> 00:45:39.430 going to take transit, are they going to take bus or rail? 00:45:39.430 --> 00:45:41.610 Fare products, are they going to go for the pass 00:45:41.610 --> 00:45:43.109 or are they going to pay as they go? 00:45:43.109 --> 00:45:45.610 And maybe there are multiple passes, which one? 00:45:45.610 --> 00:45:48.630 And then costs, not just in absolute terms. 00:45:48.630 --> 00:45:50.940 So usually, elasticity only looks 00:45:50.940 --> 00:45:54.760 at the percentage increase in fares for this mode. 00:45:54.760 --> 00:45:58.350 But what if car costs have gone up by more than that. 00:45:58.350 --> 00:46:01.570 So really, it's relative to your alternatives 00:46:01.570 --> 00:46:03.660 and each person might have a different choice. 00:46:03.660 --> 00:46:07.820 For some people, transit might be the only mode of choice 00:46:07.820 --> 00:46:10.080 [INAUDIBLE] the only mode in the choice set 00:46:10.080 --> 00:46:11.700 because they don't own a car. 00:46:11.700 --> 00:46:15.610 For others, there may be multiple options. 00:46:15.610 --> 00:46:18.741 So the other aspect of it is that exogenous factors 00:46:18.741 --> 00:46:21.240 influence demand response and they're not usually controlled 00:46:21.240 --> 00:46:21.739 for. 00:46:21.739 --> 00:46:24.420 So fuel prices, employment, residential development, 00:46:24.420 --> 00:46:27.210 tax policy, sociodemographics, new modes-- 00:46:27.210 --> 00:46:30.230 the introduction of TNCs like Uber and Lyft 00:46:30.230 --> 00:46:36.720 and other companies have potentially affected ridership, 00:46:36.720 --> 00:46:37.730 we think they have. 00:46:37.730 --> 00:46:40.250 And most of these elasticity models 00:46:40.250 --> 00:46:43.610 don't really sort of control for that 00:46:43.610 --> 00:46:46.460 to get the real fare elasticity, and that 00:46:46.460 --> 00:46:51.740 would be what you would want to do, an accurate model. 00:46:51.740 --> 00:46:52.240 OK. 00:46:52.240 --> 00:46:53.740 Communication of fare policy. 00:46:53.740 --> 00:46:56.470 So obviously, to the public it needs to be communicated. 00:46:56.470 --> 00:46:57.970 You can do that on your website, you 00:46:57.970 --> 00:47:00.890 can have near fare vending machines there's usually you 00:47:00.890 --> 00:47:03.060 know an explanation of the prices, 00:47:03.060 --> 00:47:05.179 you have a customer service booths. 00:47:05.179 --> 00:47:06.970 If there's going to be a fare change, again 00:47:06.970 --> 00:47:09.386 you post things on the website; there's flyers and posters 00:47:09.386 --> 00:47:11.740 everywhere; public hearings, of course, 00:47:11.740 --> 00:47:17.080 to sort of hear complaints and defend the hike. 00:47:17.080 --> 00:47:20.290 And I wanted to add one sort of third emerging but very 00:47:20.290 --> 00:47:25.110 important aspect that we haven't figured out, 00:47:25.110 --> 00:47:28.300 and it's via APIs or standard feeds for trip planners. 00:47:28.300 --> 00:47:32.020 So if you have an app or if you use Google Maps to get transit 00:47:32.020 --> 00:47:34.990 directions and you want it to calculate what your fare will 00:47:34.990 --> 00:47:38.290 be, unless the agency has a simple fare structure, 00:47:38.290 --> 00:47:39.082 that can't be done. 00:47:39.082 --> 00:47:41.248 Even for agencies that have a simple fare structure, 00:47:41.248 --> 00:47:43.240 most of them don't provide the two files 00:47:43.240 --> 00:47:46.600 in GTFS that allow that calculation to happen. 00:47:46.600 --> 00:47:49.410 And for agencies like London that have a capping scheme, 00:47:49.410 --> 00:47:51.820 the GTFS specification does not provide the means 00:47:51.820 --> 00:47:56.564 by which to input the fare rules that would accurately 00:47:56.564 --> 00:47:57.730 calculate the person's fare. 00:47:57.730 --> 00:47:59.229 Because you would need to keep track 00:47:59.229 --> 00:48:04.410 of how much they've taken service over a week, 00:48:04.410 --> 00:48:05.560 over a day, et cetera. 00:48:05.560 --> 00:48:08.680 So that we haven't figured out. 00:48:08.680 --> 00:48:11.500 Third party apps like the transit app 00:48:11.500 --> 00:48:14.170 would love to be able to calculate fares 00:48:14.170 --> 00:48:15.541 accurately and display them. 00:48:15.541 --> 00:48:17.290 Because they can do that for Uber and they 00:48:17.290 --> 00:48:22.665 can do that for other modes, so why not for transit as well? 00:48:22.665 --> 00:48:24.790 And although they would prefer to do that in-house, 00:48:24.790 --> 00:48:28.330 one alternative to that is to define 00:48:28.330 --> 00:48:30.550 some sort of standard API where it's 00:48:30.550 --> 00:48:34.720 up to the agency to calculate what that would be. 00:48:34.720 --> 00:48:37.150 But you make an API call and you say, 00:48:37.150 --> 00:48:39.340 this is the hypothetical trip that I'm planning. 00:48:39.340 --> 00:48:43.810 Give me how much [INAUDIBLE] and the agency would respond. 00:48:43.810 --> 00:48:47.170 But for performance reasons I think 00:48:47.170 --> 00:48:48.700 the transit app would prefer to be 00:48:48.700 --> 00:48:50.146 able to do that with their servers 00:48:50.146 --> 00:48:51.520 so that there's a quick response. 00:48:54.310 --> 00:48:57.250 Partnerships, another sort of emerging trend. 00:48:57.250 --> 00:48:59.270 So employer partnerships is one example. 00:48:59.270 --> 00:49:01.080 I'll give you the example of MIT, actually. 00:49:01.080 --> 00:49:03.910 MIT has something called AccessMyCommute. 00:49:03.910 --> 00:49:06.010 Our lab was actually involved in that. 00:49:06.010 --> 00:49:10.360 So every MIT student or employee badge 00:49:10.360 --> 00:49:13.060 has a Charlie chip embedded inside. 00:49:13.060 --> 00:49:15.610 You can look at the back and it says, Charlie inside. 00:49:15.610 --> 00:49:20.560 And to employees, it's marketed as an unlimited use pass. 00:49:20.560 --> 00:49:23.560 So I can use my MIT ID and ride the MBTA. 00:49:27.020 --> 00:49:30.220 MIT gets billed on a pay-as-you-go basis. 00:49:30.220 --> 00:49:34.780 So the reason that was the agreement 00:49:34.780 --> 00:49:38.380 is that a lot of people don't use transit enough 00:49:38.380 --> 00:49:40.280 to justify the value of a pass. 00:49:40.280 --> 00:49:43.960 So there was no way that MIT would 00:49:43.960 --> 00:49:47.080 agree to pay a pass for every employee. 00:49:47.080 --> 00:49:49.690 And it's in the interest of MIT to foment, 00:49:49.690 --> 00:49:53.500 to encourage ridership, so to decrease saliency, 00:49:53.500 --> 00:49:56.560 in this case to the extreme by making it free. 00:49:56.560 --> 00:50:01.950 So one way of doing that is to say, market it as a free pass. 00:50:01.950 --> 00:50:07.770 There is a mix of people that had passes and people that 00:50:07.770 --> 00:50:09.546 ride less often. 00:50:09.546 --> 00:50:10.920 AUDIENCE: Could MIT reach a point 00:50:10.920 --> 00:50:13.570 where it might actually want everyone to the pass 00:50:13.570 --> 00:50:16.780 without the [INAUDIBLE] employees knowing it 00:50:16.780 --> 00:50:17.280 [INAUDIBLE]? 00:50:17.280 --> 00:50:17.490 GABRIEL SANCHEZ-MARTINEZ: I don't 00:50:17.490 --> 00:50:19.780 think so because many, many employees-- yeah, 00:50:19.780 --> 00:50:20.880 it's not close. 00:50:20.880 --> 00:50:26.610 And some people ride a lot more so MIT 00:50:26.610 --> 00:50:30.150 in some ways for some employees is paying more 00:50:30.150 --> 00:50:32.070 because they're paying by ride. 00:50:32.070 --> 00:50:34.560 But many people that have a subsidized pass 00:50:34.560 --> 00:50:38.530 because their break even point is so much lower 00:50:38.530 --> 00:50:39.930 don't really break even. 00:50:39.930 --> 00:50:42.570 And so it works out to-- 00:50:42.570 --> 00:50:45.360 AUDIENCE: The next project that project might do 00:50:45.360 --> 00:50:47.940 is to find a way to predict based 00:50:47.940 --> 00:50:51.282 on the employee [INAUDIBLE] this information 00:50:51.282 --> 00:50:53.490 and they make a prediction for what he's going to do. 00:50:53.490 --> 00:50:55.460 And you might save MIT a fair amount of money. 00:50:55.460 --> 00:50:56.751 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:50:56.751 --> 00:50:58.800 Well, one of the reasons MIT decided 00:50:58.800 --> 00:51:00.870 to do this is because parking is very expensive, 00:51:00.870 --> 00:51:03.720 real state is very expensive-- very expensive. 00:51:03.720 --> 00:51:06.840 And that's, of course, not true everywhere. 00:51:06.840 --> 00:51:08.070 It's not true in most places. 00:51:08.070 --> 00:51:11.130 Most places have free parking, plenty of land. 00:51:11.130 --> 00:51:14.370 But in a place that has high density, building new parking 00:51:14.370 --> 00:51:17.190 is very expensive and you could use that real state 00:51:17.190 --> 00:51:18.220 for something else. 00:51:18.220 --> 00:51:20.850 In fact, MIT has taken a parking down near Sloan 00:51:20.850 --> 00:51:22.730 and is making a new building there. 00:51:22.730 --> 00:51:26.970 So it costs less to provide transit to people 00:51:26.970 --> 00:51:28.860 and to encourage people to take transit 00:51:28.860 --> 00:51:31.350 than to build new parking and maintain parking. 00:51:34.170 --> 00:51:35.530 Here's another example. 00:51:35.530 --> 00:51:39.000 So the Chicago Transit Authority has 00:51:39.000 --> 00:51:41.280 partnerships with other transportation providers. 00:51:41.280 --> 00:51:44.190 And they do that by providing-- 00:51:44.190 --> 00:51:47.550 there are some passes and some fare products that 00:51:47.550 --> 00:51:50.619 allow riding in PACE, which is the original bus provider; 00:51:50.619 --> 00:51:52.410 Metra, which is the commuter rail provider; 00:51:52.410 --> 00:51:55.470 and Divvy, which is the bike share. 00:51:55.470 --> 00:51:59.460 So you can buy things that-- 00:51:59.460 --> 00:52:02.490 there are some partnerships that allow some integration here. 00:52:02.490 --> 00:52:05.320 And there is a trend towards mobility as a service. 00:52:05.320 --> 00:52:08.030 So the idea of mobility as a service 00:52:08.030 --> 00:52:10.170 is that you make a monthly payment for a bundle 00:52:10.170 --> 00:52:12.850 of transportation options. 00:52:12.850 --> 00:52:13.990 Here's an example. 00:52:13.990 --> 00:52:18.000 You could have unlimited transit pass five bike rides and five 00:52:18.000 --> 00:52:18.840 TNC rides. 00:52:21.660 --> 00:52:24.840 So you provide this bundle. 00:52:24.840 --> 00:52:28.020 The person is purchasing my transportation from home 00:52:28.020 --> 00:52:30.860 to work and back on a combination of both. 00:52:30.860 --> 00:52:34.430 So if it's today it's inconvenient to take transit, 00:52:34.430 --> 00:52:39.960 that day I will use the bike share or the TNC. 00:52:39.960 --> 00:52:43.410 The word is out on whether this will work or not. 00:52:43.410 --> 00:52:45.720 Many agencies are interested; most of them 00:52:45.720 --> 00:52:47.250 don't want to talk about it. 00:52:47.250 --> 00:52:51.392 And we'll see what happens as time goes by. 00:52:51.392 --> 00:52:52.850 AUDIENCE: Is the reason that MIT is 00:52:52.850 --> 00:52:55.410 hesitant to extend the transit benefit to grad students 00:52:55.410 --> 00:52:58.680 because grad students tend not to be parkers so we 00:52:58.680 --> 00:53:02.520 don't have that bargaining? 00:53:02.520 --> 00:53:06.530 GABRIEL SANCHEZ-MARTINEZ: That is 00:53:06.530 --> 00:53:10.520 sort of a reason for why they shouldn't or wouldn't, but I'm 00:53:10.520 --> 00:53:13.010 not sure that it is the actual reason. 00:53:13.010 --> 00:53:16.460 I think this was always conceived as an employee 00:53:16.460 --> 00:53:17.800 benefits program. 00:53:17.800 --> 00:53:19.680 AUDIENCE: Aren't we like employees? 00:53:19.680 --> 00:53:21.180 GABRIEL SANCHEZ-MARTINEZ: Not quite. 00:53:21.180 --> 00:53:22.970 So yeah. 00:53:22.970 --> 00:53:26.210 So I don't really know the actual reason. 00:53:26.210 --> 00:53:28.520 Sometimes decisions are made for-- 00:53:28.520 --> 00:53:32.000 you justify them in all sorts of ways and then someone decides. 00:53:32.000 --> 00:53:36.091 So many other factors play in. 00:53:36.091 --> 00:53:36.590 OK. 00:53:36.590 --> 00:53:38.260 Fare control. 00:53:38.260 --> 00:53:40.210 Tap in, you know what that is. 00:53:40.210 --> 00:53:42.650 You interact with a fare box or fare gate 00:53:42.650 --> 00:53:45.450 only when you're entering the system or vehicle. 00:53:45.450 --> 00:53:48.980 Tap in and tap out, you do that when you enter and exit. 00:53:48.980 --> 00:53:54.650 Usually, tap in and tap out is for systems that have zones 00:53:54.650 --> 00:53:57.110 and you tap out so that we know if you've 00:53:57.110 --> 00:53:59.210 traveled multiple zones or not and therefore, we 00:53:59.210 --> 00:54:01.790 know how much to charge you because of the segmentation 00:54:01.790 --> 00:54:04.550 by geography. 00:54:04.550 --> 00:54:08.570 It also happens when there are multiple agencies sharing 00:54:08.570 --> 00:54:09.660 the revenue. 00:54:09.660 --> 00:54:18.800 So intercity bus in some cities, if a person takes a bus route 00:54:18.800 --> 00:54:22.400 within a city, then that revenue would go to that city. 00:54:22.400 --> 00:54:25.400 If the person travels across cities, 00:54:25.400 --> 00:54:27.650 then the revenue would be shared across cities. 00:54:27.650 --> 00:54:31.070 So that's one way you might want to do this. 00:54:31.070 --> 00:54:33.849 I know that it does exist in some places for bus 00:54:33.849 --> 00:54:35.390 where you have to tap in and tap out, 00:54:35.390 --> 00:54:38.940 although it's much less common for bus than for rail. 00:54:38.940 --> 00:54:40.218 Henry. 00:54:40.218 --> 00:54:45.670 AUDIENCE: If the tap system introduces a day pass, 00:54:45.670 --> 00:54:49.269 for example, where [INAUDIBLE] go to multiple-- 00:54:49.269 --> 00:54:50.560 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:54:50.560 --> 00:54:51.250 Yeah. 00:54:51.250 --> 00:54:51.720 AUDIENCE: [INAUDIBLE] 00:54:51.720 --> 00:54:53.011 GABRIEL SANCHEZ-MARTINEZ: Yeah. 00:54:53.011 --> 00:54:55.031 AUDIENCE: Then would they just [INAUDIBLE].. 00:54:55.031 --> 00:54:57.280 GABRIEL SANCHEZ-MARTINEZ: I don't know how they do it. 00:54:57.280 --> 00:54:59.530 Funding formulas are a huge debate. 00:54:59.530 --> 00:55:04.060 And as an engineer, I would propose to look at their OD 00:55:04.060 --> 00:55:09.250 and figure out the most fair way of allocating that revenue. 00:55:09.250 --> 00:55:13.747 But how would they actually do it is probably simpler. 00:55:13.747 --> 00:55:15.330 Another thing about tap in and tap out 00:55:15.330 --> 00:55:18.380 is that you may require internal vending machines. 00:55:18.380 --> 00:55:20.970 It may increase your capital costs. 00:55:20.970 --> 00:55:21.690 Why is that? 00:55:21.690 --> 00:55:24.660 Because you have multiple costs. 00:55:24.660 --> 00:55:27.260 The price of your ride depends on how far you ride 00:55:27.260 --> 00:55:28.590 and if you cross zones or not. 00:55:28.590 --> 00:55:30.131 So you might have just enough balance 00:55:30.131 --> 00:55:33.900 to enter, not enough to exit if you cross multiple zones. 00:55:33.900 --> 00:55:37.170 So you might have to top up the card 00:55:37.170 --> 00:55:38.754 before being allowed to exit. 00:55:38.754 --> 00:55:40.170 That means you need a fare vending 00:55:40.170 --> 00:55:45.165 machine that a person can use to top up 00:55:45.165 --> 00:55:47.250 if they don't have a pass. 00:55:47.250 --> 00:55:50.070 It means that you might need a station attendant 00:55:50.070 --> 00:55:51.030 to deal with somebody. 00:55:51.030 --> 00:55:53.640 If the fare vending machine is not working 00:55:53.640 --> 00:55:56.460 or my bank card is not working and is being rejected, 00:55:56.460 --> 00:55:59.910 I need to get out of the station. 00:55:59.910 --> 00:56:00.690 I'm trapped. 00:56:00.690 --> 00:56:02.430 So what do I do? 00:56:02.430 --> 00:56:05.400 So it can increase costs in that way. 00:56:05.400 --> 00:56:06.570 You had a comment? 00:56:06.570 --> 00:56:07.570 OK. 00:56:07.570 --> 00:56:08.831 No. 00:56:08.831 --> 00:56:09.330 Right. 00:56:09.330 --> 00:56:12.030 So that's one potential issue. 00:56:12.030 --> 00:56:15.850 Something that some agencies do is allowing a negative balance. 00:56:15.850 --> 00:56:18.420 So when we are softening the blows 00:56:18.420 --> 00:56:22.590 to allow when you tap out to go below zero. 00:56:22.590 --> 00:56:28.800 Next time you fill up the card, you top up, if you put in $20 00:56:28.800 --> 00:56:31.540 and you had a $1 deficit, your balance is now $19. 00:56:34.470 --> 00:56:37.920 You can't do that if cards are [INAUDIBLE] because people 00:56:37.920 --> 00:56:40.590 will then toss the card and get a new one. 00:56:40.590 --> 00:56:43.410 So you make people buy the. 00:56:43.410 --> 00:56:48.160 The cards are maybe $10 and a $5 balance or something. 00:56:48.160 --> 00:56:51.810 So there's a disincentive to tossing the card 00:56:51.810 --> 00:56:54.180 and getting a new one. 00:56:54.180 --> 00:56:56.280 And by the way, smart cards cost money. 00:56:56.280 --> 00:56:58.350 They're not that cheap. 00:56:58.350 --> 00:57:01.920 Some agencies have paid $10 per card for their card. 00:57:05.080 --> 00:57:07.410 Proof of payments, we sort of discussed that. 00:57:07.410 --> 00:57:10.840 You buy a ticket. 00:57:10.840 --> 00:57:12.840 It might be OD, it might be flat fare. 00:57:12.840 --> 00:57:15.480 It doesn't matter, you buy the ticket that you need. 00:57:15.480 --> 00:57:18.990 And then when you board, there's usually some validator 00:57:18.990 --> 00:57:21.180 that prints a time stamp. 00:57:21.180 --> 00:57:23.400 It can be electronic as well. 00:57:23.400 --> 00:57:26.220 And what it requires is fare inspection. 00:57:26.220 --> 00:57:29.070 So at random, a fare inspector will board the vehicle 00:57:29.070 --> 00:57:32.580 and have a handheld device requiring people to confirm. 00:57:32.580 --> 00:57:36.660 If it's an electronic system, then they'll tap on here 00:57:36.660 --> 00:57:39.120 and I want to make sure that you've actually validated. 00:57:39.120 --> 00:57:41.127 If it's paper, then they would look at the paper 00:57:41.127 --> 00:57:43.210 and make sure that you've canceled it or validated 00:57:43.210 --> 00:57:44.970 that paper ticket. 00:57:44.970 --> 00:57:46.350 And otherwise, big fine. 00:57:46.350 --> 00:57:51.720 So the discouragement to fare evasion is through fining 00:57:51.720 --> 00:57:53.790 and through policing. 00:57:53.790 --> 00:57:57.270 But the research shows that there are still 00:57:57.270 --> 00:57:59.127 higher fare evasion rates on systems 00:57:59.127 --> 00:58:00.210 that are proof of payment. 00:58:00.210 --> 00:58:06.798 People opportunistically board and they don't pay anything. 00:58:06.798 --> 00:58:07.770 OK. 00:58:07.770 --> 00:58:08.800 Fare technology. 00:58:08.800 --> 00:58:12.310 So we've talked about structure and policy, 00:58:12.310 --> 00:58:13.630 let's talk about technology. 00:58:13.630 --> 00:58:17.280 So this is how an AFC system is usually set up. 00:58:17.280 --> 00:58:18.780 You have a fare engine. 00:58:18.780 --> 00:58:22.530 It's a server that has all the fare rules in it 00:58:22.530 --> 00:58:27.300 so it can calculate fares based on a sequence of taps. 00:58:27.300 --> 00:58:28.950 And it's connected to different things. 00:58:28.950 --> 00:58:30.325 So a fare vending machine, that's 00:58:30.325 --> 00:58:35.130 where you purchase your card or your ticket. 00:58:35.130 --> 00:58:37.890 So these are usually fiberoptic connections 00:58:37.890 --> 00:58:44.090 or hardline connections and so they're fast. 00:58:44.090 --> 00:58:46.350 It's a real-time validation. 00:58:46.350 --> 00:58:48.810 As soon as you buy a card and add value 00:58:48.810 --> 00:58:50.820 to it from a fare vending machine, 00:58:50.820 --> 00:58:53.130 that balance is usually updated. 00:58:53.130 --> 00:58:56.430 If the fare engine has a balance in it, 00:58:56.430 --> 00:58:58.140 it will be updated in the fare engine. 00:58:58.140 --> 00:58:59.430 That's not always the case. 00:58:59.430 --> 00:59:01.620 There's many kinds of systems. 00:59:01.620 --> 00:59:04.170 So customer service booths are also connected, 00:59:04.170 --> 00:59:06.210 fare gates are usually connected as well. 00:59:06.210 --> 00:59:11.060 So when you tap into a fare gate, 00:59:11.060 --> 00:59:13.690 there's a quick communication with the fare engine saying, 00:59:13.690 --> 00:59:15.169 is this card valid, yes or no? 00:59:15.169 --> 00:59:15.960 Is there a balance? 00:59:15.960 --> 00:59:17.100 If so, deduct balance. 00:59:17.100 --> 00:59:19.470 This is how much balance needs to be deducted. 00:59:19.470 --> 00:59:20.850 And the fare gate opens. 00:59:20.850 --> 00:59:24.960 Or maybe the fare engine says, no, the card's not good. 00:59:24.960 --> 00:59:27.550 Sound the alarm, don't open. 00:59:27.550 --> 00:59:29.040 There are other ways of doing it. 00:59:29.040 --> 00:59:32.190 So the card itself might hold the balance and the fare gate 00:59:32.190 --> 00:59:35.290 might have the logic to decide whether to open or not. 00:59:35.290 --> 00:59:39.420 But there's a whole range of ways in which this 00:59:39.420 --> 00:59:40.900 can be configured. 00:59:40.900 --> 00:59:42.810 Fare boxes and validators are usually 00:59:42.810 --> 00:59:47.640 on vehicles, although there can be handheld validators. 00:59:47.640 --> 00:59:51.840 Again, a validator is more often used for proof of payment. 00:59:51.840 --> 00:59:55.530 A fare box is used when it's pay-as-you-go or pass, 00:59:55.530 --> 00:59:58.890 so that's what we have here in Boston. 00:59:58.890 --> 01:00:00.300 There's two ways of doing this. 01:00:00.300 --> 01:00:03.120 It used to be the case that they were not 01:00:03.120 --> 01:00:05.860 communicating in real time with a fare engine. 01:00:05.860 --> 01:00:08.700 So the card and the fare box have a little chat 01:00:08.700 --> 01:00:11.700 and decide how much value to deduct. 01:00:11.700 --> 01:00:15.930 Later, the bus reaches the garage or it's being cleaned 01:00:15.930 --> 01:00:18.420 and at the garage there's a garage server that 01:00:18.420 --> 01:00:20.700 wirelessly probes the fare box and extracts 01:00:20.700 --> 01:00:21.864 all the data from it. 01:00:21.864 --> 01:00:23.280 And only then will the fare engine 01:00:23.280 --> 01:00:27.120 receive the information about what transactions have 01:00:27.120 --> 01:00:28.650 been done in that fare box. 01:00:28.650 --> 01:00:33.080 So you might have, again, a real time validation. 01:00:33.080 --> 01:00:35.680 It can be a little slow over the air. 01:00:35.680 --> 01:00:40.740 So what happens, and this is being done London, for example, 01:00:40.740 --> 01:00:42.930 with contactless bank cards, what 01:00:42.930 --> 01:00:47.220 happens is that maybe the first time the card is seen, 01:00:47.220 --> 01:00:51.090 there is this dialogue and validation. 01:00:51.090 --> 01:00:55.050 After the card is marked as trusted, 01:00:55.050 --> 01:00:57.420 then it can just validate immediately 01:00:57.420 --> 01:01:02.610 and the actual deduction can wait a little bit 01:01:02.610 --> 01:01:05.250 and be sent in batches to the fare engine. 01:01:07.950 --> 01:01:10.080 There's a cash vault. That's where 01:01:10.080 --> 01:01:16.800 if your system accepts cash when the bus or fare box reaches 01:01:16.800 --> 01:01:20.010 the garage, that cash vault is taken out 01:01:20.010 --> 01:01:23.610 and the cash needs to be counted and moved out of the fare box 01:01:23.610 --> 01:01:25.950 and to a safe place. 01:01:25.950 --> 01:01:28.710 Smart cards themselves are computers. 01:01:28.710 --> 01:01:32.880 They are small computers with a battery and a clock 01:01:32.880 --> 01:01:36.185 and some amount of processing power. 01:01:36.185 --> 01:01:37.560 So they have some memory, so they 01:01:37.560 --> 01:01:41.640 can hold some history about your last trips. 01:01:41.640 --> 01:01:46.480 They can hold how much balance there is in some cases. 01:01:46.480 --> 01:01:50.430 They are able to encrypt and decrypt communication 01:01:50.430 --> 01:01:54.600 so that it's harder to break the security 01:01:54.600 --> 01:01:57.510 and say add balance your card. 01:01:57.510 --> 01:02:00.360 And I want to add although this is not necessarily 01:02:00.360 --> 01:02:03.150 part of the fare system, data [INAUDIBLE].. 01:02:03.150 --> 01:02:05.610 Increasingly, we are doing a lot of analytics. 01:02:05.610 --> 01:02:07.650 So usually, there's some connection 01:02:07.650 --> 01:02:11.070 between the live production system and a data warehouse 01:02:11.070 --> 01:02:14.080 where on a daily or a weekly basis, 01:02:14.080 --> 01:02:17.170 there's an extract of data into a data warehouse. 01:02:17.170 --> 01:02:23.790 And that allows analysts to run heavy queries on a server that 01:02:23.790 --> 01:02:26.900 is dedicated to analysis and not on the live system [INAUDIBLE] 01:02:26.900 --> 01:02:29.658 whether or not to open a fare, if that makes sense. 01:02:33.465 --> 01:02:33.965 OK. 01:02:33.965 --> 01:02:35.020 Fare media. 01:02:35.020 --> 01:02:37.720 Any questions about fare technology and the framework 01:02:37.720 --> 01:02:38.945 of how this is set up? 01:02:38.945 --> 01:02:40.570 There are lots of variants and I didn't 01:02:40.570 --> 01:02:45.150 get into the details of how any one system works. 01:02:45.150 --> 01:02:45.930 OK. 01:02:45.930 --> 01:02:47.220 So fare media. 01:02:47.220 --> 01:02:51.450 Cash, we all know what that is; tokens. 01:02:51.450 --> 01:02:55.020 So I don't need to go into how cash 01:02:55.020 --> 01:02:57.300 is minted or anything, right? 01:02:57.300 --> 01:03:02.460 So tokens are little coins that you buy with money 01:03:02.460 --> 01:03:05.200 and it's like a ticket in some ways. 01:03:05.200 --> 01:03:07.290 It has some value and sometimes it's 01:03:07.290 --> 01:03:10.530 possible to be able to have to pay two tokens for a ride, 01:03:10.530 --> 01:03:13.039 although usually it's just one. 01:03:13.039 --> 01:03:14.580 They're being phased out because they 01:03:14.580 --> 01:03:16.950 can print paper tickets with magnetic stripes 01:03:16.950 --> 01:03:18.350 and there's more data to it. 01:03:21.056 --> 01:03:21.960 AUDIENCE: [INAUDIBLE] 01:03:21.960 --> 01:03:23.251 GABRIEL SANCHEZ-MARTINEZ: Yeah. 01:03:23.251 --> 01:03:24.530 So they still exist. 01:03:24.530 --> 01:03:25.620 AUDIENCE: [INAUDIBLE] 01:03:25.620 --> 01:03:26.911 GABRIEL SANCHEZ-MARTINEZ: Yeah. 01:03:26.911 --> 01:03:29.050 They still exist. 01:03:29.050 --> 01:03:30.140 Yeah. 01:03:30.140 --> 01:03:33.298 This is a Boston token, by the way, this one right here. 01:03:33.298 --> 01:03:35.440 AUDIENCE: [INAUDIBLE] 01:03:35.440 --> 01:03:37.640 GABRIEL SANCHEZ-MARTINEZ: No. 01:03:37.640 --> 01:03:40.070 I'm pretty sure this is an old Boston token. 01:03:40.070 --> 01:03:40.570 Yeah. 01:03:40.570 --> 01:03:41.880 AUDIENCE: [INAUDIBLE] 01:03:41.880 --> 01:03:43.171 GABRIEL SANCHEZ-MARTINEZ: Yeah. 01:03:43.171 --> 01:03:43.940 Yeah. 01:03:43.940 --> 01:03:45.870 Maybe I'm wrong, but I think that's the case. 01:03:45.870 --> 01:03:46.370 OK. 01:03:46.370 --> 01:03:46.910 Tickets. 01:03:46.910 --> 01:03:50.030 So tickets are they're printed on paper 01:03:50.030 --> 01:03:51.380 and they have a magnetic stripe. 01:03:51.380 --> 01:03:53.710 They don't have a computer in them. 01:03:53.710 --> 01:03:55.700 The magnetic stripe holds information 01:03:55.700 --> 01:03:59.550 that can be encrypted. 01:03:59.550 --> 01:04:02.390 So it can be, but it's not always encrypted. 01:04:02.390 --> 01:04:08.000 It's a little easier for there to be fraud with tickets. 01:04:08.000 --> 01:04:12.167 So if there are questions on tickets, 01:04:12.167 --> 01:04:13.250 I'll talk more about them. 01:04:13.250 --> 01:04:15.530 Smart cards I've talked about. 01:04:15.530 --> 01:04:17.030 Mobile ticketing is another option. 01:04:17.030 --> 01:04:19.340 So that's a recent development. 01:04:19.340 --> 01:04:26.690 You have an app, you pay for a ticket for your ride 01:04:26.690 --> 01:04:27.470 on the app. 01:04:27.470 --> 01:04:30.560 And it's usually kind of a proof of payment system 01:04:30.560 --> 01:04:35.630 where a fare inspector comes along and scans a QR code 01:04:35.630 --> 01:04:39.770 on the app, proving that you've purchased a valid fare 01:04:39.770 --> 01:04:42.140 for that OD pair, for example. 01:04:42.140 --> 01:04:44.600 This is being used here in commuter rail in Boston. 01:04:47.580 --> 01:04:51.980 And then the sort of newest form is contactless bank cards. 01:04:51.980 --> 01:04:58.330 So that's a wireless Visa, MasterCard, credit charge card, 01:04:58.330 --> 01:05:03.060 debit card and you tap it on the-- 01:05:03.060 --> 01:05:04.470 there you go. 01:05:04.470 --> 01:05:07.610 You can pass it if you want. 01:05:07.610 --> 01:05:08.190 Can I see it? 01:05:12.360 --> 01:05:14.160 So you tap it just like you would 01:05:14.160 --> 01:05:17.550 a Charlie card or an Oyster card in London 01:05:17.550 --> 01:05:24.450 and there's the dialogue between the fare and the card. 01:05:24.450 --> 01:05:26.950 Strong encryption. 01:05:26.950 --> 01:05:29.640 As I said, the full validation process 01:05:29.640 --> 01:05:33.120 doesn't always happen, especially in bus, 01:05:33.120 --> 01:05:34.864 because you can trust the card. 01:05:34.864 --> 01:05:36.030 There is a separate antenna. 01:05:36.030 --> 01:05:38.820 Of course, the frequencies and the protocols are different. 01:05:38.820 --> 01:05:40.380 But there are good standards for this 01:05:40.380 --> 01:05:43.890 and London has moved to this. 01:05:43.890 --> 01:05:46.860 CTA's system is based on this technology, 01:05:46.860 --> 01:05:54.890 even though they opted to only use token cards. 01:05:54.890 --> 01:05:56.500 So it's the same technology, but it's 01:05:56.500 --> 01:05:58.070 branded as their own thing. 01:05:58.070 --> 01:06:00.850 And I don't know if they can-- 01:06:00.850 --> 01:06:01.810 now it does, right? 01:06:01.810 --> 01:06:03.229 But at the beginning, it wasn't. 01:06:03.229 --> 01:06:05.020 AUDIENCE: I met someone who had a problem-- 01:06:05.020 --> 01:06:06.311 GABRIEL SANCHEZ-MARTINEZ: Yeah. 01:06:06.311 --> 01:06:07.630 AUDIENCE: [INAUDIBLE] 01:06:07.630 --> 01:06:08.921 GABRIEL SANCHEZ-MARTINEZ: Yeah. 01:06:08.921 --> 01:06:14.170 So with this, as you have more and more cards 01:06:14.170 --> 01:06:15.670 on your wallet that are contactless, 01:06:15.670 --> 01:06:19.030 you should get like a copper-lined wallet 01:06:19.030 --> 01:06:21.360 or something, to like copper sleeves. 01:06:21.360 --> 01:06:26.480 They exist to sort of prevent two reads from the same card. 01:06:26.480 --> 01:06:29.450 AUDIENCE: So if I have a contactless bank card, 01:06:29.450 --> 01:06:31.940 can I just walk up to one of those systems and tap in? 01:06:31.940 --> 01:06:33.340 GABRIEL SANCHEZ-MARTINEZ: Yeah. 01:06:33.340 --> 01:06:34.460 AUDIENCE: [INAUDIBLE] absolutely nothing? 01:06:34.460 --> 01:06:35.380 GABRIEL SANCHEZ-MARTINEZ: Absolutely nothing. 01:06:35.380 --> 01:06:37.090 You just get on the bus. 01:06:37.090 --> 01:06:38.290 Yeah. 01:06:38.290 --> 01:06:39.860 It used to be the case in Chicago 01:06:39.860 --> 01:06:41.350 that it was kind of their own thing 01:06:41.350 --> 01:06:43.656 and you were given a card. 01:06:43.656 --> 01:06:45.280 It was the same thing as a credit card, 01:06:45.280 --> 01:06:49.720 but just as a token with no balance or account on it. 01:06:49.720 --> 01:06:53.530 And now they've enabled this. 01:06:53.530 --> 01:06:55.461 So that's the new way. 01:06:55.461 --> 01:06:56.960 AUDIENCE: But there are other issues 01:06:56.960 --> 01:07:00.940 about that because there are issues of what 01:07:00.940 --> 01:07:04.360 if the card is from a different country and things like that? 01:07:04.360 --> 01:07:06.485 GABRIEL SANCHEZ-MARTINEZ: Again, you do validation. 01:07:06.485 --> 01:07:07.390 There are policies. 01:07:07.390 --> 01:07:10.019 That's what these card companies do. 01:07:10.019 --> 01:07:10.810 They're good at it. 01:07:10.810 --> 01:07:13.690 There is insurance for fraud. 01:07:13.690 --> 01:07:17.240 So they-- yeah. 01:07:17.240 --> 01:07:17.870 Yeah. 01:07:17.870 --> 01:07:19.630 AUDIENCE: It's a big push from Visa and MasterCard 01:07:19.630 --> 01:07:22.180 to get [INAUDIBLE] contact list because [INAUDIBLE] money 01:07:22.180 --> 01:07:23.440 [INAUDIBLE]. 01:07:23.440 --> 01:07:26.110 So every way they can make it easier to spend money 01:07:26.110 --> 01:07:27.140 is better for them. 01:07:27.140 --> 01:07:28.080 GABRIEL SANCHEZ-MARTINEZ: Eli, you have a comment? 01:07:28.080 --> 01:07:30.430 AUDIENCE: And in Chicago, it doesn't process transfers 01:07:30.430 --> 01:07:30.970 or passes. 01:07:30.970 --> 01:07:34.851 They're just charging you a la carte, one trip. 01:07:34.851 --> 01:07:37.690 GABRIEL SANCHEZ-MARTINEZ: So there's 01:07:37.690 --> 01:07:39.190 they don't have a capping scheme. 01:07:39.190 --> 01:07:40.735 AUDIENCE: Right. 01:07:40.735 --> 01:07:41.860 And it doesn't [INAUDIBLE]. 01:07:41.860 --> 01:07:43.600 GABRIEL SANCHEZ-MARTINEZ: So if you want a pass, 01:07:43.600 --> 01:07:45.350 that's not an option for you is what you're saying. 01:07:45.350 --> 01:07:47.010 AUDIENCE: You need to actually [INAUDIBLE].. 01:07:47.010 --> 01:07:47.340 GABRIEL SANCHEZ-MARTINEZ: OK. 01:07:47.340 --> 01:07:47.840 Yeah. 01:07:47.840 --> 01:07:51.010 So that's an interesting comment which no one heard. 01:07:55.360 --> 01:07:58.440 I'm sure the comments on the back are also interesting. 01:07:58.440 --> 01:08:01.390 AUDIENCE: They're just asking me about [INAUDIBLE].. 01:08:01.390 --> 01:08:04.960 GABRIEL SANCHEZ-MARTINEZ: I'm generally interested. 01:08:04.960 --> 01:08:06.410 Can you share with the class? 01:08:06.410 --> 01:08:08.576 AUDIENCE: They were just asking whether you actually 01:08:08.576 --> 01:08:09.602 have to [INAUDIBLE]. 01:08:09.602 --> 01:08:10.810 GABRIEL SANCHEZ-MARTINEZ: Oh. 01:08:10.810 --> 01:08:11.820 OK. 01:08:11.820 --> 01:08:13.736 AUDIENCE: You do have to take it [INAUDIBLE].. 01:08:13.736 --> 01:08:16.000 GABRIEL SANCHEZ-MARTINEZ: As opposed to EMV, yeah. 01:08:16.000 --> 01:08:17.375 AUDIENCE: But [INAUDIBLE] Chicago 01:08:17.375 --> 01:08:19.670 have it because I thought it [INAUDIBLE] 01:08:19.670 --> 01:08:21.340 penetration [INAUDIBLE]. 01:08:21.340 --> 01:08:23.715 GABRIEL SANCHEZ-MARTINEZ: But I think they will increase. 01:08:23.715 --> 01:08:26.380 And that I think could be driven by agencies in big cities, 01:08:26.380 --> 01:08:26.899 actually. 01:08:26.899 --> 01:08:29.050 That's one example of that. 01:08:29.050 --> 01:08:31.600 The US has been a little late in adopting credit card 01:08:31.600 --> 01:08:32.560 technology in general. 01:08:35.851 --> 01:08:36.350 All right. 01:08:36.350 --> 01:08:37.069 Smart cards. 01:08:37.069 --> 01:08:41.270 So I've already talked a little bit about them. 01:08:41.270 --> 01:08:43.819 They enable more complex fare structures, 01:08:43.819 --> 01:08:47.060 they allow faster boarding and higher gatebank throughput 01:08:47.060 --> 01:08:49.520 because the validations are much faster than a ticket. 01:08:49.520 --> 01:08:52.520 A magnetic stripe ticket needs to be fed in 01:08:52.520 --> 01:08:54.439 and read multiple times sometimes 01:08:54.439 --> 01:08:56.479 and it takes more time to do that. 01:08:56.479 --> 01:08:59.330 Certainly, cash is slower. 01:08:59.330 --> 01:09:03.020 So account registration is one of the benefits of smart cards. 01:09:03.020 --> 01:09:07.040 You don't have to do this, but you often can. 01:09:07.040 --> 01:09:09.620 So you open an account and that gives you balance protection. 01:09:09.620 --> 01:09:12.350 If your card is lost, you can log into your account 01:09:12.350 --> 01:09:13.880 and remove that card. 01:09:13.880 --> 01:09:16.011 The balance could be retained on the account 01:09:16.011 --> 01:09:18.260 so you can get a new card associated with that account 01:09:18.260 --> 01:09:20.220 and you have your balance. 01:09:20.220 --> 01:09:22.340 Autoload is another convenience. 01:09:22.340 --> 01:09:26.090 You might have a rule that says, if my balance drops below $5, 01:09:26.090 --> 01:09:27.500 charge $20. 01:09:27.500 --> 01:09:31.700 So if it's top up or pay-as-you-go, 01:09:31.700 --> 01:09:35.180 then this one feature. 01:09:35.180 --> 01:09:40.939 Of course, it's much better for us sort of researchers and data 01:09:40.939 --> 01:09:43.430 nerds, better data for analysis and planning, 01:09:43.430 --> 01:09:45.020 longitudinal studies. 01:09:45.020 --> 01:09:46.970 You can track a token longitudinally 01:09:46.970 --> 01:09:49.970 and understand more about behavior. 01:09:49.970 --> 01:09:53.859 It's embeddable in employee and student budgets, as we've seen. 01:09:53.859 --> 01:09:55.400 They are more expensive than tickets. 01:09:55.400 --> 01:09:58.520 Now there are some smart tickets so you can print a card 01:09:58.520 --> 01:10:00.800 with a chip circuit inside. 01:10:00.800 --> 01:10:02.660 Those are a little cheaper. 01:10:02.660 --> 01:10:06.110 One issue with smart cards is that a lot of these systems 01:10:06.110 --> 01:10:07.370 are proprietary. 01:10:07.370 --> 01:10:09.830 And there are some standards, but there are several 01:10:09.830 --> 01:10:11.970 standards-- too many of them-- 01:10:11.970 --> 01:10:14.080 and they don't all work everywhere 01:10:14.080 --> 01:10:16.670 and so you can't really use a card 01:10:16.670 --> 01:10:18.470 from one system on another system. 01:10:18.470 --> 01:10:20.130 That's usually not the case. 01:10:20.130 --> 01:10:21.770 There are some exceptions to that. 01:10:21.770 --> 01:10:25.280 There is a move towards open source hardware specifications. 01:10:25.280 --> 01:10:27.000 New York is a good example of that. 01:10:27.000 --> 01:10:28.970 They recently did some procurements 01:10:28.970 --> 01:10:32.670 where they asked all the parts to be-- 01:10:32.670 --> 01:10:35.120 it was for an AVL system in this case-- 01:10:35.120 --> 01:10:40.380 to be open source hardware and communications technology. 01:10:40.380 --> 01:10:42.720 That means they're not locked into that vendor. 01:10:42.720 --> 01:10:47.780 So if they need to replace one aspect of that system, 01:10:47.780 --> 01:10:50.090 they can hire someone else that also 01:10:50.090 --> 01:10:52.760 complies with that standard and that part can be replaced. 01:10:52.760 --> 01:10:56.270 That lowers costs, increases competition, 01:10:56.270 --> 01:10:58.855 and it means again that they're not locked into that vendor. 01:10:58.855 --> 01:10:59.354 Lee. 01:10:59.354 --> 01:11:01.904 AUDIENCE: [INAUDIBLE] security at all or [INAUDIBLE].. 01:11:01.904 --> 01:11:03.820 GABRIEL SANCHEZ-MARTINEZ: Usually, the systems 01:11:03.820 --> 01:11:05.990 have strong standard encryption. 01:11:05.990 --> 01:11:08.180 Part of the standard is what encryption is used 01:11:08.180 --> 01:11:14.480 and what encryption algorithms are used, hashing, et cetera. 01:11:14.480 --> 01:11:14.990 Yeah. 01:11:14.990 --> 01:11:17.840 So the Charlie card has pretty weak encryption, 01:11:17.840 --> 01:11:25.390 40 bits I think, and the same key on every card. 01:11:25.390 --> 01:11:28.930 So a long time ago, someone cracked the key by brute force 01:11:28.930 --> 01:11:31.000 and that's the same key on every Charlie card. 01:11:31.000 --> 01:11:34.300 So once that key is public, anyone with the equipment 01:11:34.300 --> 01:11:38.080 could load a fake balance to the card 01:11:38.080 --> 01:11:41.770 because the cards are the ones having the balance in the AFC 01:11:41.770 --> 01:11:44.470 system here and not the fare engine. 01:11:44.470 --> 01:11:47.290 That means that effectively, you can lie to the fare system 01:11:47.290 --> 01:11:50.620 and say that you have more money than you have. 01:11:50.620 --> 01:11:53.380 Much harder to do with any modern system, 01:11:53.380 --> 01:11:55.210 certainly contact with a bank card 01:11:55.210 --> 01:11:57.730 which has a very strong encryption and unique keys 01:11:57.730 --> 01:11:58.909 per card. 01:11:58.909 --> 01:12:00.950 AUDIENCE: Did MIT students get caught doing that? 01:12:00.950 --> 01:12:02.700 GABRIEL SANCHEZ-MARTINEZ: The MIT students 01:12:02.700 --> 01:12:04.540 were the first to do this. 01:12:04.540 --> 01:12:08.120 They were caught and there was some issue. 01:12:08.120 --> 01:12:10.650 Which I think it ended up being just a slap on the wrist, 01:12:10.650 --> 01:12:15.557 but there was a lot of issue with it. 01:12:15.557 --> 01:12:17.140 Because they wanted to publish a paper 01:12:17.140 --> 01:12:20.000 or publish something and then that was an issue. 01:12:20.000 --> 01:12:20.500 OK. 01:12:20.500 --> 01:12:24.410 So integration across agencies of a region is possible. 01:12:24.410 --> 01:12:28.160 That's one of the reasons for moving towards smart cards, 01:12:28.160 --> 01:12:31.030 especially in regions that have many providers. 01:12:31.030 --> 01:12:33.100 But it's challenging, and usually the challenges 01:12:33.100 --> 01:12:37.190 are less technical and more about trust and business 01:12:37.190 --> 01:12:40.420 and who gets to decide what color the card is 01:12:40.420 --> 01:12:44.314 or exactly what the fare policy will be. 01:12:44.314 --> 01:12:45.730 This is another interesting thing. 01:12:45.730 --> 01:12:49.540 In some countries, and I'm thinking specifically 01:12:49.540 --> 01:12:53.980 of Japan and China, there are examples of retail stores 01:12:53.980 --> 01:12:55.960 that have partnered with a transit agency 01:12:55.960 --> 01:13:00.707 and you can use the card at retail locations. 01:13:00.707 --> 01:13:02.040 Hong Kong is an example of that. 01:13:02.040 --> 01:13:06.880 There's more than 100 retail locations that accept Octopus. 01:13:06.880 --> 01:13:10.570 And in Japan, I know that there are entire cities where 01:13:10.570 --> 01:13:13.600 there's like a whole mall and this is what most people use 01:13:13.600 --> 01:13:15.160 as a charge card. 01:13:15.160 --> 01:13:18.490 So that's interesting. 01:13:18.490 --> 01:13:21.100 If you don't have very tough privacy laws, 01:13:21.100 --> 01:13:23.780 as you do in Europe and the US-- 01:13:23.780 --> 01:13:25.750 I'm not sure what they are in Japan, 01:13:25.750 --> 01:13:28.420 but there was some research where 01:13:28.420 --> 01:13:33.130 they were combining all the data from shopping and transit 01:13:33.130 --> 01:13:38.110 to get really interesting behavioral analytics. 01:13:38.110 --> 01:13:39.790 And you can do interesting things 01:13:39.790 --> 01:13:46.290 like offer discounts at stores if you do something in transit. 01:13:46.290 --> 01:13:51.295 So there's a lot of very interesting [INAUDIBLE].. 01:13:51.295 --> 01:13:53.346 AUDIENCE: Our company did the same thing 01:13:53.346 --> 01:13:55.796 and once [INAUDIBLE]. 01:14:00.075 --> 01:14:04.264 So after that, we are very careful about dealing 01:14:04.264 --> 01:14:05.169 with [INAUDIBLE]. 01:14:05.169 --> 01:14:06.460 GABRIEL SANCHEZ-MARTINEZ: Yeah. 01:14:06.460 --> 01:14:07.680 Yeah. 01:14:07.680 --> 01:14:10.200 Thanks for sharing. 01:14:10.200 --> 01:14:11.400 So contactless bank cards. 01:14:11.400 --> 01:14:12.899 We started talking about this, we've 01:14:12.899 --> 01:14:14.430 mentioned that multiple times. 01:14:14.430 --> 01:14:16.560 Here's the logic for it. 01:14:16.560 --> 01:14:18.090 A transit agency would prefer not 01:14:18.090 --> 01:14:20.550 having to deal with the complexities and costs of fare 01:14:20.550 --> 01:14:21.776 collection. 01:14:21.776 --> 01:14:23.400 Essentially, if you have a fare system, 01:14:23.400 --> 01:14:26.320 it's like having a small bank. 01:14:26.320 --> 01:14:28.890 It doesn't give out loans, it doesn't do other things 01:14:28.890 --> 01:14:30.780 that banks do, but you are dealing 01:14:30.780 --> 01:14:33.570 with money and the cards and customer service related 01:14:33.570 --> 01:14:36.150 to cards and all those things. 01:14:36.150 --> 01:14:37.950 Banks do that, and they do it very well. 01:14:37.950 --> 01:14:42.840 So why not outsource that part of the business to banks? 01:14:42.840 --> 01:14:46.780 There are multiple banks so they can get competition. 01:14:46.780 --> 01:14:50.190 So credit card companies specialize in payment; 01:14:50.190 --> 01:14:52.380 contactless bank cards are very secure, 01:14:52.380 --> 01:14:55.950 much more so than say the Charlie card; 01:14:55.950 --> 01:14:59.860 you can use cards like this in several ways. 01:14:59.860 --> 01:15:03.150 So you could directly use them for payment 01:15:03.150 --> 01:15:05.070 as you would at a retail location 01:15:05.070 --> 01:15:09.570 or you can use the token that identifies that card, tying it 01:15:09.570 --> 01:15:10.890 to an account. 01:15:10.890 --> 01:15:14.460 And then your fare engine can calculate what the fare will be 01:15:14.460 --> 01:15:15.990 and capping and all that. 01:15:15.990 --> 01:15:21.270 And then on a regular basis, there 01:15:21.270 --> 01:15:25.050 would be a charge to your credit card or account 01:15:25.050 --> 01:15:29.940 for whatever total was calculated. 01:15:29.940 --> 01:15:34.140 So that also enables Apple Pay and Android Pay. 01:15:34.140 --> 01:15:36.580 If you can put your credit card on your phone, 01:15:36.580 --> 01:15:40.350 then you can use your phone to get into the system. 01:15:40.350 --> 01:15:43.110 In London, people love that. 01:15:43.110 --> 01:15:44.860 It reduces fare collection costs. 01:15:44.860 --> 01:15:47.610 So why? 01:15:47.610 --> 01:15:49.770 Not necessarily obvious. 01:15:49.770 --> 01:15:52.050 So it simplifies customer communication. 01:15:52.050 --> 01:15:55.260 You don't have to issue new cards to people, 01:15:55.260 --> 01:15:59.310 you don't have to explain how you pay because people 01:15:59.310 --> 01:16:01.500 know how to use cards. 01:16:01.500 --> 01:16:03.150 It relies on open standards so there's 01:16:03.150 --> 01:16:04.660 more competition on the market. 01:16:04.660 --> 01:16:06.780 There's several banks, for example, 01:16:06.780 --> 01:16:10.980 and you could have banks compete for who 01:16:10.980 --> 01:16:16.140 will give you a better deal in terms of fees on these cards. 01:16:19.080 --> 01:16:21.480 It outsources some aspects of customer service to banks. 01:16:21.480 --> 01:16:23.340 So if my card stopped working, who's 01:16:23.340 --> 01:16:27.630 going to get the phone call, the transit agency or the bank? 01:16:27.630 --> 01:16:30.570 And it eliminates costs of sort of creating and distributing 01:16:30.570 --> 01:16:31.380 smart cards. 01:16:31.380 --> 01:16:35.250 So there's several ways in which this is reducing costs. 01:16:35.250 --> 01:16:39.580 London has successfully been able to reduce 01:16:39.580 --> 01:16:46.950 the number of station attendants and especially of booths 01:16:46.950 --> 01:16:49.580 where they sell Oyster cards and have customer service related 01:16:49.580 --> 01:16:50.540 to Oyster cards. 01:16:50.540 --> 01:16:53.726 They've been able to close those and save money. 01:16:53.726 --> 01:16:55.100 And then there's an equity issue, 01:16:55.100 --> 01:16:57.420 and this is what is always sort of brought up 01:16:57.420 --> 01:16:59.750 when you talk about contactless bank cards, which 01:16:59.750 --> 01:17:01.580 is what about the unbanked? 01:17:01.580 --> 01:17:06.120 The poorest people are the ones who don't have a bank account. 01:17:06.120 --> 01:17:08.120 So what can you do about that? 01:17:08.120 --> 01:17:10.340 Does that make this not work? 01:17:10.340 --> 01:17:13.160 So there are multiple ways of addressing that. 01:17:13.160 --> 01:17:16.520 An agency can issue cards that have preloaded balance. 01:17:16.520 --> 01:17:20.510 So that person could interact with the agency, not the bank. 01:17:20.510 --> 01:17:23.030 They don't necessarily need a bank account. 01:17:23.030 --> 01:17:25.240 You can have banks offer free accounts. 01:17:25.240 --> 01:17:26.780 So you can try to encourage people 01:17:26.780 --> 01:17:30.500 to open accounts that have no fee, no minimum balance. 01:17:30.500 --> 01:17:34.830 So some [INAUDIBLE] special accounts created for that. 01:17:34.830 --> 01:17:36.800 And one other issue is that you must 01:17:36.800 --> 01:17:40.310 be able to obtain these cards at many locations. 01:17:40.310 --> 01:17:43.520 If you move towards a system like this 01:17:43.520 --> 01:17:47.630 and you can only get one of these sort of equity 01:17:47.630 --> 01:17:54.380 issued cards if you want from one station downtown, that 01:17:54.380 --> 01:17:58.580 means that the person who depends on two or three buses 01:17:58.580 --> 01:18:02.450 to get into downtown can't easily get a card. 01:18:02.450 --> 01:18:05.600 So you must be able to maybe get these 01:18:05.600 --> 01:18:08.390 at several retail locations or have 01:18:08.390 --> 01:18:13.980 fare vending machines in multiple places, 01:18:13.980 --> 01:18:17.510 especially in zones of high concentrations of people 01:18:17.510 --> 01:18:22.960 that depend on transit and don't have bank accounts. 01:18:22.960 --> 01:18:23.860 OK. 01:18:23.860 --> 01:18:27.800 So there are ways of dealing with that. 01:18:27.800 --> 01:18:31.119 Boston is considering now what they call AFC 2.0. 01:18:31.119 --> 01:18:32.660 They're sort of branding it that way. 01:18:32.660 --> 01:18:37.100 And they're probably going to be moving forward 01:18:37.100 --> 01:18:41.840 with the London model of using contactless bank cards. 01:18:41.840 --> 01:18:43.910 So we'll see how it plays out and we'll 01:18:43.910 --> 01:18:47.210 see what that does to the payment industry in general 01:18:47.210 --> 01:18:52.590 in the US in terms of that means more people will have these. 01:18:52.590 --> 01:18:57.710 Does that mean that a lot of readers at retail locations 01:18:57.710 --> 01:19:01.050 will start accepting that, and what will that do? 01:19:01.050 --> 01:19:01.910 It's a question. 01:19:01.910 --> 01:19:04.620 We'll see what happens. 01:19:04.620 --> 01:19:05.310 OK. 01:19:05.310 --> 01:19:06.650 Last topic is fare analytics. 01:19:06.650 --> 01:19:09.440 So obviously, lots of data is generated by these systems. 01:19:09.440 --> 01:19:11.120 We've used it for all the [INAUDIBLE],, 01:19:11.120 --> 01:19:13.380 you've looked at it in your homework. 01:19:13.380 --> 01:19:18.950 So here's an example from one of the students in the transit lab 01:19:18.950 --> 01:19:20.510 who has one more year to graduate, 01:19:20.510 --> 01:19:22.940 but he's done great work. 01:19:22.940 --> 01:19:24.980 His name is Andrew Stutz. 01:19:24.980 --> 01:19:27.260 And one of the examples of his work 01:19:27.260 --> 01:19:33.710 is dashboard, where you can look at all sorts of fare 01:19:33.710 --> 01:19:41.750 transactions by fare product, by pass or no pass, by user type. 01:19:41.750 --> 01:19:44.750 And you can sort of filter and see exactly who's 01:19:44.750 --> 01:19:46.430 using your system how often. 01:19:46.430 --> 01:19:48.292 Let me see if I can show that to you live. 01:19:51.370 --> 01:19:51.870 OK. 01:19:51.870 --> 01:19:57.840 So here's AFC validations for October 2015. 01:19:57.840 --> 01:20:02.736 So if I want to see, for example, 01:20:02.736 --> 01:20:04.110 here we're looking at everything. 01:20:04.110 --> 01:20:07.080 So we're looking at tickets and cards and passes and monthly 01:20:07.080 --> 01:20:08.830 passes and pay-as-you-go. 01:20:08.830 --> 01:20:12.502 So if I want to look at stored value, so no passes, 01:20:12.502 --> 01:20:13.710 I'll sort of filter that out. 01:20:13.710 --> 01:20:17.640 Then of course tariff type becomes 100% pay-as-you-go. 01:20:17.640 --> 01:20:20.340 And then you see the breakdown between cards and tickets 01:20:20.340 --> 01:20:22.260 for stored value use. 01:20:22.260 --> 01:20:27.120 We see the trends by time of day here and by date. 01:20:27.120 --> 01:20:31.331 So you can see that weekends are lower and the user type 01:20:31.331 --> 01:20:31.830 right here. 01:20:31.830 --> 01:20:36.630 So adults very high and then senior students below that. 01:20:36.630 --> 01:20:42.670 If you want to look at passes, for example, 01:20:42.670 --> 01:20:46.540 we could then say passes and monthly passes in particular. 01:20:46.540 --> 01:20:49.320 You see that most of those are cards, not tickets. 01:20:49.320 --> 01:20:51.360 Makes sense. 01:20:51.360 --> 01:20:55.244 Higher number of students than we saw before, et cetera. 01:20:55.244 --> 01:20:56.910 So you can really sort of play with this 01:20:56.910 --> 01:21:02.510 and understand how the market segments itself 01:21:02.510 --> 01:21:06.960 with these different fare products 01:21:06.960 --> 01:21:09.150 that people are purchasing based on the fare policy. 01:21:09.150 --> 01:21:14.160 And there's another dashboard for fare revenue. 01:21:14.160 --> 01:21:19.860 So these are fare revenues by month across many years. 01:21:19.860 --> 01:21:23.607 And you can see the total up here. 01:21:23.607 --> 01:21:24.940 But you can break it down again. 01:21:24.940 --> 01:21:28.750 So you can look at how much of it is monthly passes 01:21:28.750 --> 01:21:31.980 and you can see how much of it is normal commuter 01:21:31.980 --> 01:21:33.540 pass versus commuter rail. 01:21:33.540 --> 01:21:39.990 And you can see, if you're interested in students, 01:21:39.990 --> 01:21:41.790 here's your students. 01:21:41.790 --> 01:21:44.910 So you can really slice and dice it any way you want to. 01:21:44.910 --> 01:21:46.410 And there's many more applications, 01:21:46.410 --> 01:21:49.811 but I just wanted to demonstrate a few. 01:21:49.811 --> 01:21:50.310 OK. 01:21:50.310 --> 01:21:52.120 So that was it. 01:21:52.120 --> 01:21:55.980 I'm not going to lecture anyone for this semester. 01:21:55.980 --> 01:21:59.820 So I'll still come. 01:21:59.820 --> 01:22:02.700 And I'll let you know about the test 01:22:02.700 --> 01:22:05.140 as soon as I have more information on it. 01:22:05.140 --> 01:22:09.040 Any questions on fare policy, fare structure, 01:22:09.040 --> 01:22:10.300 fare technology? 01:22:10.300 --> 01:22:10.800 No. 01:22:10.800 --> 01:22:11.860 All right. 01:22:11.860 --> 01:22:12.360 Thank you. 01:22:12.360 --> 01:22:15.110 [APPLAUSE]