WEBVTT

1
00:00:00.388 --> 00:00:03.555
[Light music]

2
00:00:05.580 --> 00:00:07.260
- Did I drive for Uber?

3
00:00:07.260 --> 00:00:08.400
In fact, I did.

4
00:00:08.400 --> 00:00:11.910
I enjoyed driving for
Uber for a few weeks.

5
00:00:11.910 --> 00:00:15.300
It was the summer of 2016,
and I did that

6
00:00:15.300 --> 00:00:18.210
because I was doing
some research on Uber,

7
00:00:18.210 --> 00:00:20.760
and I thought I need to know
what the world looks like

8
00:00:20.760 --> 00:00:21.720
to an Uber driver.

9
00:00:21.720 --> 00:00:23.070
I also drove for Lyft.

10
00:00:23.070 --> 00:00:25.080
I very much enjoyed
driving for Uber.

11
00:00:25.080 --> 00:00:26.400
I didn't make a lot of money.

12
00:00:26.400 --> 00:00:29.520
I definitely didn't make enough
money as an Uber driver

13
00:00:29.520 --> 00:00:34.200
to... you know, be tempted
to quit my day job at MIT.

14
00:00:34.200 --> 00:00:35.580
I viewed Uber driving

15
00:00:35.580 --> 00:00:39.090
as an opportunity to collect
data on labor market outcomes.

16
00:00:39.090 --> 00:00:40.890
I enjoyed the experience.

17
00:00:40.890 --> 00:00:42.840
I enjoyed talking to passengers.

18
00:00:42.840 --> 00:00:45.180
I don't know if my passengers
enjoyed talking to me.

19
00:00:45.180 --> 00:00:47.280
So I would ask passengers
where they worked,

20
00:00:47.280 --> 00:00:50.550
and whether they liked their
jobs and how much they make

21
00:00:50.550 --> 00:00:53.730
and had they ever thought
about driving for Uber.

22
00:00:53.730 --> 00:00:56.820
And some people liked
that sort of discussion

23
00:00:56.820 --> 00:00:59.670
but many people did
not like that too much.

24
00:00:59.670 --> 00:01:01.263
So my ratings weren't too high.

25
00:01:02.100 --> 00:01:04.950
The research study that I did,
it's a paper I wrote

26
00:01:04.950 --> 00:01:07.050
with Jonathan Hall
who works at Uber

27
00:01:07.050 --> 00:01:10.980
and Sydney Caldwell who was
my PhD student at the time.

28
00:01:10.980 --> 00:01:14.610
We were interested in aspects
of the driver labor market.

29
00:01:14.610 --> 00:01:16.500
One question that's very important

30
00:01:16.500 --> 00:01:19.530
to Uber and very important
in labor economics

31
00:01:19.530 --> 00:01:22.320
is how elastic drivers are.

32
00:01:22.320 --> 00:01:24.450
How responsive they are to wages.

33
00:01:24.450 --> 00:01:26.820
Do they come out more on more days

34
00:01:26.820 --> 00:01:30.780
and drive for more hours
when pay is high.

35
00:01:30.780 --> 00:01:33.360
Academic literature and
labor economics

36
00:01:33.360 --> 00:01:34.830
has a vigorous debate between

37
00:01:34.830 --> 00:01:39.180
a sort of a more behavioral
view where people drive less

38
00:01:39.180 --> 00:01:40.350
when their wages are high,

39
00:01:40.350 --> 00:01:42.570
because they have a kind of a target

40
00:01:42.570 --> 00:01:44.850
and when they hit their target,
they go home.

41
00:01:44.850 --> 00:01:46.890
That's very hard to rationalize

42
00:01:46.890 --> 00:01:50.160
with a straightforward
intertemporal optimization model.

43
00:01:50.160 --> 00:01:53.280
In a straightforward intertemporal
optimization model

44
00:01:53.280 --> 00:01:55.290
people should drive
more when wages are high.

45
00:01:55.290 --> 00:01:58.050
You should take advantage
of the low-hanging fruit.

46
00:01:58.050 --> 00:02:01.020
So I wanted to test
the alternative views

47
00:02:01.020 --> 00:02:02.463
of driver labor supply.

48
00:02:03.510 --> 00:02:05.358
That was one goal of
the project, and indeed

49
00:02:05.358 --> 00:02:07.710
drivers are highly elastic.

50
00:02:07.710 --> 00:02:10.410
They drive more when
we raise their wages

51
00:02:10.410 --> 00:02:11.970
as part of the treatment.

52
00:02:11.970 --> 00:02:14.250
The other thing that
Sydney and Jonathan and I

53
00:02:14.250 --> 00:02:17.700
were interested in is...
is the nature of the contract

54
00:02:17.700 --> 00:02:19.230
that Uber drivers work under.

55
00:02:19.230 --> 00:02:21.960
So what is the sort of
essential difference

56
00:02:21.960 --> 00:02:25.950
between driving for Uber and
driving, say, a taxi cab?

57
00:02:25.950 --> 00:02:29.010
And the core economics of the models

58
00:02:29.010 --> 00:02:30.420
from the point of view of the drivers

59
00:02:30.420 --> 00:02:33.180
comes down to the compensation scheme.

60
00:02:33.180 --> 00:02:35.880
The typical sort of yellow cab driver

61
00:02:35.880 --> 00:02:37.770
has to lease the right to drive.

62
00:02:37.770 --> 00:02:38.760
The right to drive

63
00:02:38.760 --> 00:02:41.190
is embodied in something
called a medallion.

64
00:02:41.190 --> 00:02:43.200
It's actually a piece
of medal on the vehicle.

65
00:02:43.200 --> 00:02:45.870
The medal is not worth
much, but the right to drive

66
00:02:45.870 --> 00:02:47.640
used to be worth quite a lot.

67
00:02:47.640 --> 00:02:49.980
Say in New York City
a medallion might cost

68
00:02:49.980 --> 00:02:53.100
a million dollars and most
drivers can't afford that.

69
00:02:53.100 --> 00:02:55.380
So they lease a medallion
for somebody.

70
00:02:55.380 --> 00:02:58.800
So they might pay 700 or
a thousand dollars a week

71
00:02:58.800 --> 00:02:59.820
for the right to drive,

72
00:02:59.820 --> 00:03:01.860
but then they keep
every dollar they earn.

73
00:03:01.860 --> 00:03:03.090
Uber doesn't work that way.

74
00:03:03.090 --> 00:03:05.700
Uber just takes
a percentage of your fare.

75
00:03:05.700 --> 00:03:09.240
So you don't have to put any
money upfront to drive, say,

76
00:03:09.240 --> 00:03:11.760
today or this week for Uber.

77
00:03:11.760 --> 00:03:13.800
Now the economics
of that are interesting

78
00:03:13.800 --> 00:03:15.990
because somebody who's
gonna drive long hours

79
00:03:15.990 --> 00:03:18.240
should actually prefer
the medallion model.

80
00:03:18.240 --> 00:03:20.460
Whereas, if you're really
just driving part-time,

81
00:03:20.460 --> 00:03:24.090
you'd like that Uber model
where they're taking a fee.

82
00:03:24.090 --> 00:03:26.700
We tested the... the driver's views

83
00:03:26.700 --> 00:03:29.100
of those compensation
schemes by offering

84
00:03:29.100 --> 00:03:31.860
some Uber drivers the
opportunity to lease.

85
00:03:31.860 --> 00:03:34.320
And we found that very
few drivers like leasing,

86
00:03:34.320 --> 00:03:36.093
they much prefer the Uber model.

87
00:03:36.720 --> 00:03:38.090
[light music]

88
00:03:38.130 --> 00:03:40.440
- Ready to master econometrics?

89
00:03:40.440 --> 00:03:43.170
Click here to embark on
an educational journey

90
00:03:43.170 --> 00:03:47.580
with Joshua Angrist, aka Master Joshway.

91
00:03:47.580 --> 00:03:49.200
Or if you'd like to watch more

92
00:03:49.200 --> 00:03:51.660
from this interview series, click here.