0:00:00.400,0:00:03.050
- [Woman] The path[br]from cause to effect

0:00:03.050,0:00:05.539
is dark and dangerous,

0:00:05.539,0:00:08.900
but the weapons[br]of econometrics are strong,

0:00:09.211,0:00:13.700
wield differences-in-differences[br]when witnessing parallel trends.

0:00:14.450,0:00:16.850
♪ [music] ♪

0:00:20.100,0:00:21.423
Masters of metrics

0:00:21.423,0:00:24.800
look for convincing[br]ceteris paribus comparisons.

0:00:25.100,0:00:29.419
The ideal comparison contrasts[br]treatment and control groups

0:00:29.419,0:00:30.600
that look similar.

0:00:30.600,0:00:34.700
But sometimes this sort[br]of comparability is elusive.

0:00:34.700,0:00:36.805
When treatment and control groups

0:00:36.805,0:00:40.100
evolve similarly[br]in the absence of treatment,

0:00:40.100,0:00:42.307
even if from different[br]starting points,

0:00:42.307,0:00:44.900
there's hope for causal inference.

0:00:45.600,0:00:48.400
The weapon that exploits[br]parallel evolution,

0:00:48.664,0:00:50.886
masters say parallel trends,

0:00:50.886,0:00:53.273
is called differences-in-differences...

0:00:53.273,0:00:54.400
- [Man whispering][br]Differences-in-differences

0:00:54.400,0:00:56.900
- ...or DD for short.[br]- [Man] Alright.

0:00:56.900,0:00:59.987
Let's see how DD[br]can help us understand

0:00:59.987,0:01:02.888
one of the most important[br]economic events

0:01:02.888,0:01:04.370
in US history.

0:01:05.300,0:01:08.300
- [Joshua] Look back with me now[br]at the Great Depression --

0:01:08.800,0:01:12.200
the worst economic catastrophe[br]our country has ever known.

0:01:13.100,0:01:16.200
Unemployment hit 25% in 1933 --

0:01:16.600,0:01:19.114
a level not seen before or since.

0:01:19.473,0:01:22.100
Millions lost their homes[br]or their land.

0:01:22.600,0:01:24.737
Suicide spiked, and hungry families

0:01:24.737,0:01:26.766
relied on soup kitchens[br]and bread lines

0:01:26.766,0:01:28.155
to keep from starving.

0:01:29.400,0:01:34.000
- Economists argue fiercely over[br]the causes of the Great Depression.

0:01:34.000,0:01:36.983
Most agree, however,[br]that a key piece of the puzzle

0:01:36.983,0:01:39.458
is an epidemic of bank failures.

0:01:39.800,0:01:41.900
This was before deposit insurance.

0:01:42.100,0:01:46.553
So if your bank went bankrupt,[br]your savings disappeared with it.

0:01:46.553,0:01:47.672
- [Cashier] Closing your account?

0:01:47.672,0:01:48.892
- [Customer] Yes, sir.[br]I'm closing my account.

0:01:48.892,0:01:50.726
- I wouldn't leave a nickel[br]in this bank.

0:01:52.600,0:01:56.100
- Faced with a banking crisis,[br]the Central Bank has a choice:

0:01:56.400,0:01:58.524
lend freely to troubled banks

0:01:58.524,0:02:01.100
or stand aside and refuse to lend.

0:02:01.500,0:02:05.440
Lending freely to banks in trouble[br]is called "easy money."

0:02:05.440,0:02:08.100
Refusing to lend[br]is called "tight money."

0:02:10.200,0:02:12.872
- [Joshua] Monetarist masters[br]Milton Friedman and Anna Schwartz

0:02:12.872,0:02:14.882
famously called[br]the Great Depression

0:02:14.882,0:02:16.350
the "Great Contraction,"

0:02:16.800,0:02:18.262
accusing the Federal Reserve

0:02:18.262,0:02:21.200
of inflicting a misguided policy[br]of tight money

0:02:21.200,0:02:24.000
on the nation's teetering[br]financial institutions.

0:02:24.400,0:02:25.873
They argued that easy money

0:02:25.873,0:02:27.985
would have kept[br]many banks in business,

0:02:27.985,0:02:29.700
shortening the Great Depression.

0:02:30.400,0:02:32.239
But others disagree!

0:02:32.239,0:02:33.769
If banks are insolvent

0:02:33.769,0:02:35.954
because of unwise[br]lending decisions,

0:02:35.954,0:02:38.900
then bailouts just encourage[br]more foolishness.

0:02:39.600,0:02:42.965
Economists called this problem[br]"moral hazard."

0:02:42.965,0:02:46.100
The debate over bailouts[br]in moral hazard continues today.

0:02:46.500,0:02:48.599
Should financial behemoth[br]Lehman Brothers

0:02:48.599,0:02:51.500
had been allowed to fail[br]on the eve of the Great Recession,

0:02:52.000,0:02:54.703
in an ideal world,[br]we'd answer this question

0:02:54.703,0:02:58.400
by applying different Fed policies[br]to randomly selected regions.

0:02:59.000,0:03:00.250
But we can still learn a lot

0:03:00.250,0:03:02.119
by using differences-in-differences

0:03:02.119,0:03:06.300
to compare trends across areas[br]with different monetary policies.

0:03:10.900,0:03:12.522
- [Woman 2] How's that even possible?

0:03:12.522,0:03:15.623
Don't the same Fed policies[br]apply to all banks in the US?

0:03:15.623,0:03:17.400
- [Man] Yeah.[br]- Good question.

0:03:17.700,0:03:21.484
The Federal Reserve System[br]is divided into 12 districts,

0:03:21.484,0:03:23.860
each headed by a regional bank.

0:03:24.301,0:03:27.467
Today, Fed policy is set[br]at the national level.

0:03:27.467,0:03:31.973
But in the 1930s, regional Feds[br]could do pretty much as they liked.

0:03:31.973,0:03:33.276
- [Man] Ah, interesting.

0:03:33.276,0:03:35.500
- And here's what's[br]so awesome about that.

0:03:35.500,0:03:39.302
In 1930, the Atlanta Fed,[br]running the 6th district,

0:03:39.302,0:03:41.473
followed an easy money policy,

0:03:41.473,0:03:45.400
sending wheelbarrows of cash[br]to rescue insolvent institutions.

0:03:45.900,0:03:48.816
The St. Louis Fed,[br]running the 8th district,

0:03:48.816,0:03:50.668
followed a tight money policy.

0:03:51.100,0:03:53.900
"Let fail the foolish!"[br]they said in St. Louis.

0:03:54.300,0:03:58.225
And so a natural experiment[br]in monetary policy was born.

0:03:58.701,0:04:02.000
Even better, this is[br]a within-state experiment.

0:04:02.000,0:04:04.347
The border between the 6th[br]and the 8th districts

0:04:04.347,0:04:06.762
ran smack through[br]the middle of Mississippi.

0:04:07.300,0:04:09.321
So northern Mississippi[br]had tight money,

0:04:09.321,0:04:11.870
while southern Mississippi[br]had easy money,

0:04:11.870,0:04:15.200
but under the same state laws[br]and banking regulations in both.

0:04:15.690,0:04:16.853
- [Woman] The treatment group

0:04:16.853,0:04:19.985
is the district 6th part[br]of Mississippi,

0:04:19.985,0:04:23.100
which had access to easy money[br]during the crisis.

0:04:23.800,0:04:25.091
The control group

0:04:25.091,0:04:27.800
is the district 8th part[br]of Mississippi,

0:04:27.800,0:04:30.225
which had tight money[br]during the crisis.

0:04:31.300,0:04:34.040
The key year[br]in our natural experiment

0:04:34.040,0:04:35.400
was 1930.

0:04:35.800,0:04:37.439
Caldwell & Company,

0:04:37.439,0:04:40.377
a massive financial empire[br]in the South

0:04:40.377,0:04:41.987
came crashing down.

0:04:42.700,0:04:46.000
Banking is a business[br]built on confidence and trust.

0:04:46.500,0:04:49.151
The Caldwell meltdown[br]caused a panic

0:04:49.151,0:04:53.049
that led to a widespread[br]bank run all at once.

0:04:53.049,0:04:55.114
Depositors wanted their money back,

0:04:55.114,0:04:58.200
causing banks to go bankrupt[br]and shut their doors.

0:05:01.000,0:05:03.321
We'll use differences-in-differences

0:05:03.321,0:05:06.614
to measure the effect[br]of contrasting monetary policies

0:05:06.614,0:05:09.164
in response to the Caldwell crisis.

0:05:12.500,0:05:16.279
This figure plots the number[br]of banks in Mississippi by year,

0:05:16.279,0:05:18.762
for the 8th and 6th districts.

0:05:19.273,0:05:24.200
Let's start in 1929 -- a year[br]before the Caldwell crash.

0:05:24.200,0:05:27.646
There are 169 banks[br]open in the 8th,

0:05:27.646,0:05:31.242
and 141 banks open in the 6th.

0:05:31.242,0:05:32.574
Over the next year,

0:05:32.574,0:05:37.000
we see a similar handful[br]of banks fail, in both districts.

0:05:37.400,0:05:40.135
The change in the number[br]of banks in operation

0:05:40.135,0:05:42.168
is remarkably similar.

0:05:42.168,0:05:44.600
That's what parallel trends look like.

0:05:45.500,0:05:48.783
In November 1930, Caldwell crashes,

0:05:48.783,0:05:50.400
and the panic begins.

0:05:51.200,0:05:53.917
Banks failed frequently[br]in the 8th district,

0:05:53.917,0:05:55.300
which had tight money.

0:05:55.800,0:05:58.712
But the decline is slower[br]in the 6th district,

0:05:58.712,0:06:00.247
which had easy money.

0:06:00.800,0:06:03.000
The diverging trends in this period

0:06:03.000,0:06:06.602
might be attributable[br]to easy versus tight money.

0:06:06.602,0:06:12.020
In July of 1931, the 8th district[br]abandons tight money,

0:06:12.020,0:06:14.420
so now both districts are easy.

0:06:14.700,0:06:16.900
Parallel trends are restored.

0:06:17.300,0:06:19.229
In a counterfactual world,

0:06:19.229,0:06:22.420
where the 6th district[br]follows a tight money policy,

0:06:22.420,0:06:23.800
what might have happened?

0:06:24.201,0:06:28.547
If we extrapolate the trend[br]of the 8th district to the 6th,

0:06:28.547,0:06:29.900
it would look like this.

0:06:30.290,0:06:33.095
So the treatment-effective[br]easy money

0:06:33.095,0:06:36.462
is how much the 6th district[br]deviated from the path

0:06:36.462,0:06:38.900
implied by the 8th district trend.

0:06:41.244,0:06:44.475
How many banks[br]did the easy money treatment save?

0:06:44.475,0:06:49.021
This table reports data[br]for the treatment group, district 6,

0:06:49.021,0:06:50.162
in the first row,

0:06:50.162,0:06:54.237
and data for the control group,[br]district 8, in the second row.

0:06:54.237,0:06:57.494
The first column shows[br]the number of banks in business

0:06:57.494,0:07:00.613
before the crisis began in 1930.

0:07:00.613,0:07:03.673
The second column shows 1931.

0:07:03.903,0:07:05.607
This is the key period

0:07:05.607,0:07:08.663
when each district[br]had differing monetary policies

0:07:08.663,0:07:10.237
during the crisis.

0:07:10.237,0:07:14.347
The rightmost column[br]reports changes within the district.

0:07:14.347,0:07:20.456
District 6 lost 14 banks,[br]while district 8 lost 33.

0:07:20.456,0:07:23.973
The mathematical formula[br]for the treatment effect is simple.

0:07:24.400,0:07:28.700
We subtract the change in banks[br]in operation in the 8th district

0:07:28.900,0:07:32.200
from the change in banks[br]in operation in the 6th.

0:07:32.800,0:07:35.700
Hence, the name[br]differences-in-differences.

0:07:37.000,0:07:41.736
-14 minus -33 equals 19.

0:07:42.500,0:07:46.800
We estimate that 19 banks[br]were saved by easy money.

0:07:47.300,0:07:50.730
In practice, tables and figures[br]like those shown here

0:07:50.730,0:07:52.912
are the beginning[br]rather than the end

0:07:52.912,0:07:54.453
of a DD analysis.

0:07:55.400,0:07:57.100
The problem of how to gauge

0:07:57.100,0:08:00.167
the statistical significance[br]of DD estimates

0:08:00.167,0:08:02.394
turns out to be exceedingly tricky,

0:08:02.400,0:08:05.600
and a regression is typically[br]part of the solution.

0:08:09.200,0:08:12.336
The key assumption[br]behind a valid DD analysis

0:08:12.336,0:08:14.894
is that of parallel trends.

0:08:15.369,0:08:17.842
Recall the principle[br]of ceteris paribus --

0:08:17.842,0:08:21.700
our ideal comparison would have[br]the two districts experience

0:08:21.700,0:08:24.023
an identical business environment,

0:08:24.023,0:08:25.997
except for one factor:

0:08:25.997,0:08:27.916
easy or tight money.

0:08:29.200,0:08:32.475
Both districts would have[br]identical types of customers

0:08:32.475,0:08:35.300
who would go bankrupt[br]at exactly the same rate.

0:08:35.700,0:08:38.600
The skill of their employees[br]would be equal, and so on.

0:08:39.200,0:08:43.400
Perfect ceteris paribus comparisons[br]would allow us to clearly see

0:08:43.600,0:08:46.594
the causal effect[br]of different Fed policies.

0:08:46.594,0:08:49.019
In this case, that's not possible.

0:08:49.019,0:08:53.600
But the idea of parallel trends[br]is based on a similar concept.

0:08:53.600,0:08:57.364
If we see that the two regions[br]experienced similar trends

0:08:57.364,0:08:59.814
in the number of banks over time,

0:08:59.814,0:09:01.294
in the absence of treatment,

0:09:01.294,0:09:04.407
we can assume[br]they are good comparisons.

0:09:04.407,0:09:07.434
We see that the two districts[br]move in parallel,

0:09:07.434,0:09:09.800
both before the crisis and after.

0:09:10.100,0:09:12.400
when they have the same Fed policy.

0:09:13.100,0:09:16.028
The only time the districts[br]behaved differently

0:09:16.028,0:09:18.300
is when the Fed policy is different.

0:09:19.400,0:09:20.599
In view of this,

0:09:20.599,0:09:24.268
Fed policy is a likely cause[br]of diverging trends

0:09:24.268,0:09:26.700
from 1930 to 1931.

0:09:27.800,0:09:30.106
But we should also check[br]for other changes

0:09:30.106,0:09:31.703
unique to northern Mississippi.

0:09:31.703,0:09:33.200
- [Man] Huh?[br]- What do you mean?

0:09:33.500,0:09:35.396
- [Woman] Imagine that bad tornadoes

0:09:35.396,0:09:39.100
hit northern but not[br]southern Mississippi in 1930.

0:09:39.600,0:09:41.950
These tornadoes devastate farms,

0:09:41.950,0:09:44.300
causing farmers[br]to default on loans,

0:09:44.551,0:09:46.800
which drives their banks[br]out of business.

0:09:47.400,0:09:49.438
Then the 6th and 8th districts

0:09:49.438,0:09:52.272
would differ in not one[br]but two ways:

0:09:52.700,0:09:55.047
Fed policy and weather.

0:09:55.047,0:09:58.219
And we'd have trouble[br]identifying Fed policy

0:09:58.219,0:10:01.590
as the causal factor[br]behind increased bank failures

0:10:01.590,0:10:02.600
in the 8th.

0:10:02.600,0:10:04.248
- [Man] Ceteris is not paribus.

0:10:07.200,0:10:09.014
- DD credibility lives or dies

0:10:09.014,0:10:10.980
with the claim that the only reason

0:10:10.980,0:10:13.794
northern Mississippi[br]was special in 1930

0:10:13.794,0:10:16.161
is differing regional Fed policy.

0:10:16.600,0:10:20.530
We're in DD heaven with strong,[br]visual evidence of parallel trend.

0:10:21.277,0:10:25.549
- In general, the first step[br]in evaluating whether to use DD

0:10:25.549,0:10:30.200
is usually this type of[br]visual confirmation of parallel trends

0:10:30.200,0:10:31.700
outside of the period,

0:10:31.700,0:10:35.094
when we expect to see[br]a treatment effect.

0:10:35.094,0:10:36.985
The treatment in our example

0:10:36.985,0:10:39.835
is easy money[br]in the face of bank failures.

0:10:40.500,0:10:45.000
Metrics masters use DD[br]to explore effects of many policies,

0:10:45.800,0:10:47.900
like the minimum legal drinking age,

0:10:48.500,0:10:52.200
and environmental changes,[br]like access to clean water.

0:10:52.867,0:10:54.200
In our next video,

0:10:54.500,0:10:57.035
we'll see an example[br]of how regression is used

0:10:57.035,0:10:59.200
to implement a DD approach.

0:11:00.800,0:11:02.183
- [Narrator] Are you a teacher?

0:11:02.183,0:11:05.788
Click to explore ways[br]to use these videos in class.

0:11:05.788,0:11:08.847
If you're a learner,[br]make sure this video sticks

0:11:08.847,0:11:11.200
by taking a few quick[br]practice questions.

0:11:11.600,0:11:14.200
Or if you're ready.[br]click for the next video.

0:11:14.600,0:11:17.093
You can also check out[br]MRU's website

0:11:17.093,0:11:20.193
for more courses,[br]teacher resources, and more.

0:11:20.193,0:11:21.692
♪ [music] ♪