Introduction to Differences-in-Differences
-
0:00 - 0:06The path from cause to effect
is dark and dangerous, -
0:06 - 0:09but the weapons
of econometrics are strong, -
0:10 - 0:14wield differences-in-differences
when witnessing parallel trends. -
0:14 - 0:17♪ [music] ♪
-
0:20 - 0:21Masters of metrics
-
0:21 - 0:25look for convincing
ceteris paribus comparisons. -
0:25 - 0:29The ideal comparison contrasts
treatment and control groups -
0:29 - 0:31that look similar.
-
0:31 - 0:35But sometimes this sort
of comparability is elusive. -
0:35 - 0:37When treatment and control groups
-
0:37 - 0:40evolve similarly
in the absence of treatment, -
0:40 - 0:42even if from different
starting points, -
0:42 - 0:45there's hope for causal inference.
-
0:46 - 0:48The weapon that exploits
parallel evolution, -
0:49 - 0:51masters say parallel trends,
-
0:51 - 0:53is called differences-in-differences...
-
0:53 - 0:54(voice whispering)
Differences-in-differences -
0:54 - 0:57- ...or DD for short.
- Alright. Nice. -
0:57 - 1:00Let's see how DD
can help us understand -
1:00 - 1:03one of the most important
economic events -
1:03 - 1:04in US history.
-
1:05 - 1:08Look back with me now
at the Great Depression-- -
1:09 - 1:12the worst economic catastrophe,
our country has ever known. -
1:13 - 1:16Unemployment hit 25% in 1933--
-
1:17 - 1:20a level not seen before or since.
-
1:20 - 1:22Millions lost their homes
or their land. -
1:23 - 1:25Suicide spiked, and hungry families
-
1:25 - 1:27relied on soup kitchens
and bread lines -
1:27 - 1:28to keep from starving.
-
1:29 - 1:34Economists argue fiercely over
the causes of the Great Depression. -
1:34 - 1:37Most agree, however,
that a key piece of the puzzle -
1:37 - 1:40is an epidemic of bank failures.
-
1:40 - 1:42This was before deposit insurance.
-
1:42 - 1:46So if your bank went bankrupt,
your savings disappeared with it, -
1:53 - 1:56Faced with a banking crisis,
the central bank has a choice: -
1:56 - 1:59lend freely to troubled banks
-
1:59 - 2:01or stand aside and refuse to lend.
-
2:02 - 2:05Lending freely to banks in trouble
is called easy money. -
2:05 - 2:08Refusing to lend is called tight money.
-
2:10 - 2:13Monetarist masters Milton Friedman
and Anna Schwartz -
2:13 - 2:15famously called
the Great Depression -
2:15 - 2:16the "Great Contraction,"
-
2:17 - 2:18accusing the Federal Reserve
-
2:18 - 2:21of inflicting a misguided policy
of tight money -
2:21 - 2:24on the nation's teetering
financial institutions. -
2:24 - 2:26They argued that easy money
-
2:26 - 2:28would have kept
many banks in business, -
2:28 - 2:30shortening the Great Depression,
-
2:30 - 2:32But others disagree!
-
2:32 - 2:34If banks are insolvent
-
2:34 - 2:36because of unwise
lending decisions, -
2:36 - 2:39then bailouts just encourage
more foolishness. -
2:40 - 2:43Economists call this problem
"moral hazard." -
2:43 - 2:46The debate over bailouts
in moral hazard continues today. -
2:46 - 2:49Should financial behemoth
Lehman Brothers -
2:49 - 2:52had been allowed to fail
on the eve of the Great Recession, -
2:52 - 2:55in an ideal world,
we'd answer this question -
2:55 - 2:58applying different Fed policies
to randomly selected regions. -
2:59 - 3:00But we can still learn a lot
-
3:00 - 3:02by using differences-in-differences
-
3:02 - 3:06to compare trends across areas
with different monetary policies. -
3:11 - 3:13How's that even possible?
-
3:13 - 3:16Don't the same Fed policies
apply to all banks in the US? -
3:16 - 3:17- Yeah.
- Good question. -
3:18 - 3:21The Federal Reserve System
is divided into 12 districts, -
3:21 - 3:24each headed by a regional bank.
-
3:24 - 3:27Today, Fed policy is set
at the national level. -
3:27 - 3:32But in the 1930s, regional Feds
could do pretty much as they liked. -
3:32 - 3:33Ah, interesting.
-
3:33 - 3:36And here's what's
so awesome about that. -
3:36 - 3:39In 1930, the Atlanta Fed,
running the 6th District, -
3:39 - 3:41followed an easy money policy,
-
3:41 - 3:45sending wheelbarrows of cash
to rescue insolvent institutions, -
3:46 - 3:49The St. Louis Fed,
running the 8th District, -
3:49 - 3:51followed a tight money policy.
-
3:51 - 3:54"Let fail the foolish!"
they said in St. Louis. -
3:54 - 3:59And so a natural experiment
in monetary policy was born. -
3:59 - 4:02Even better, this is
a within-state experiment. -
4:02 - 4:04The border between the 6th
and the 8th districts -
4:04 - 4:07ran smack through
the middle of Mississippi. -
4:07 - 4:09So northern Mississippi
had tight money, -
4:09 - 4:12while southern Mississippi
had easy money, -
4:12 - 4:15but under the same state laws
in banking regulations in both. -
4:16 - 4:17The treatment group
-
4:17 - 4:20is the District 6th part
of Mississippi, -
4:20 - 4:23which had access to easy money
during the crisis. -
4:24 - 4:25The control group
-
4:25 - 4:28is the District 8th part
of Mississippi, -
4:28 - 4:30which had tight money
during the crisis. -
4:31 - 4:34The key year
in our natural experiment -
4:34 - 4:35was 1930,
-
4:36 - 4:38Caldwell & Company,
-
4:38 - 4:40a massive financial empire
in the South -
4:40 - 4:42came crashing down.
-
4:43 - 4:46Banking is a business
built on confidence and trust. -
4:46 - 4:49The Caldwell meltdown
caused a panic -
4:49 - 4:53that led to a widespread
bank run all at once. -
4:53 - 4:55Depositors wanted their money back,
-
4:55 - 4:58causing banks to go bankrupt
and shut their doors. -
5:01 - 5:03We'll use differences-in-differences
-
5:03 - 5:07to measure the effect
of contrasting monetary policies -
5:07 - 5:09in response to the Caldwell crisis.
-
5:12 - 5:16This figure plots the number
of banks in Mississippi by year, -
5:16 - 5:19for the 8th and 6th districts.
-
5:19 - 5:24Let's start in 1929 - a year
before the Caldwell crash. -
5:24 - 5:28There are 169 banks
open in the 8th, -
5:28 - 5:31and 141 banks open in the 6th.
-
5:31 - 5:33Over the next year,
-
5:33 - 5:37we see a similar handful
of banks fail, in both districts. -
5:37 - 5:40The change in the number
of banks in operation -
5:40 - 5:42is remarkably similar.
-
5:42 - 5:45That's what parallel trends look like.
-
5:46 - 5:49In November 1930, Caldwell crashes,
-
5:49 - 5:50and the panic begins.
-
5:51 - 5:53Banks fail frequently in the 8th
district, -
5:53 - 5:55which had Tight money,
-
5:56 - 6:00but the decline is slower in the
6th District which had easy money. -
6:01 - 6:03The diverging Trends in this period,
-
6:03 - 6:06might be attributable to
easy versus Tight money -
6:07 - 6:12in July, of 1931, the 8th
District, abandons type money. -
6:12 - 6:14So now both districts are easy.
-
6:15 - 6:17Parallel Trends are restored
-
6:17 - 6:22in a counterfactual world where the sixth
district follows a Tight money policy. -
6:22 - 6:24What might have happened
-
6:24 - 6:28if we extrapolate the trend of
the 8th District to the sixth, -
6:28 - 6:30it would look like this.
-
6:30 - 6:31So
-
6:31 - 6:36The treatment effective easy money is
how much the 6th District deviated From -
6:36 - 6:39the Path implied by
the 8th District trend. -
6:41 - 6:44How many banks did the
Easy Money treatment save -
6:44 - 6:50this table reports data for the treatment
group District Six in the first row -
6:50 - 6:54and data for the control group
District 8 in the second row. -
6:54 - 6:57The First Column shows the
number of banks in business, -
6:57 - 7:03before The Crisis began in 1930.
The second column shows 1931. -
7:03 - 7:09This is the key period when each district
had differing monetary policies during -
7:09 - 7:11the crisis the rightmost.
-
7:11 - 7:12Column
-
7:12 - 7:18reports changes within the district
district, 6 lost 14 Banks while District -
7:18 - 7:248 lost 33, the mathematical formula
for the treatment effect is simple. -
7:24 - 7:29We subtract the change in banks
in operation, in the 8th District -
7:29 - 7:32from the change in banks
in operation in the sixth. -
7:33 - 7:36Hence. The name
differences in differences, -
7:37 - 7:41negative 14, minus negative, 33 equals 9.
-
7:41 - 7:4219.
-
7:42 - 7:47We estimate that 19 Banks
were saved by easy money -
7:47 - 7:52in practice tables and figures like
those shown here are the beginning -
7:52 - 7:54rather than the end of a DD analysis,
-
7:55 - 8:00the problem of how to gauge the
statistical, significance of DD, estimates -
8:00 - 8:02turns out to be exceedingly, tricky
-
8:02 - 8:06and a regression is typically
part of the solution, -
8:09 - 8:11the key assumption behind a
-
8:11 - 8:18DD analysis is that of parallel Trends
recall, the principle of ceteris paribus, -
8:18 - 8:22our ideal comparison would have
the two districts experienced, -
8:22 - 8:28an identical business environment, except
for one factor easy or tight money. -
8:29 - 8:34Both districts would have identical
types of customers who would go bankrupt -
8:34 - 8:35at exactly the same rate.
-
8:36 - 8:39The skill of their employees
would be equal and so on -
8:39 - 8:43perfect ceteris, paribus comparisons
would allow us to clearly see -
8:44 - 8:49the causal effect of different fed
policies in this case, that's not possible. -
8:49 - 8:54But the idea of parallel Trends
is based on a similar concept. -
8:54 - 8:58If we see that the two regions
experienced similar Trends in the number -
8:58 - 8:59of banks over.
-
8:59 - 9:04Time in the absence of treatment, we
can assume they are good comparisons. -
9:04 - 9:10We see that the two districts move in
parallel both before the crisis. And after, -
9:10 - 9:12when they have the same Fed policy,
-
9:13 - 9:18the only time the district's behave differently
is when the Fed policy is different. -
9:19 - 9:27In view of this Fed policy is a likely
cause of diverging Trends from 1930 to 1931. -
9:28 - 9:32But we should also check for other
changes unique to northern, Mississippi. -
9:32 - 9:33What do you mean?
-
9:34 - 9:39Imagine that bad tornadoes? Hit Northern,
but not Southern. Mississippi in 1930, -
9:40 - 9:44these tornadoes devastate Farms
causing Farmers to default on loans, -
9:45 - 9:47which drives their Banks out of business.
-
9:47 - 9:52Then the sixth and eighth districts
would differ in not one, but two ways -
9:53 - 9:57Fed policy, and whether and
we'd have trouble identifying -
9:58 - 10:02Policy as the causal Factor behind
increased bank failures in the eighth. -
10:07 - 10:10DD credibility lives or
dies with the claim that the -
10:10 - 10:14only reason northern
Mississippi was special in 1930 -
10:14 - 10:16is differing. Regional Fed policy.
-
10:17 - 10:22We're in DD heaven with strong Visual
Evidence of parallel Trend in general. -
10:22 - 10:24The first step in evaluating whether
-
10:24 - 10:30to use DD is usually this type of
visual confirmation of parallel Trends -
10:30 - 10:32outside of the period.
-
10:32 - 10:37When we expect to see a treatment
effect, the treatment in our example, -
10:37 - 10:40Easy money in the face of bank failures
-
10:40 - 10:45metrics Masters, use DD to
explore effects of many policies -
10:46 - 10:48like the minimum, legal drinking age,
-
10:48 - 10:52and environmental changes
like access to clean water. -
10:53 - 10:54In our next video.
-
10:54 - 10:59We'll see an example of how regression
is used to implement a DD approach. -
11:01 - 11:07Are you a teacher click to explore ways to
use these videos in class if your learner -
11:07 - 11:11Make sure this video sticks by taking
a few quick practice questions, -
11:12 - 11:14or if you're ready.
Click for the next video. -
11:15 - 11:20You can also check out Mr. Use website for
more courses, teacher resources and more.
- Title:
- Introduction to Differences-in-Differences
- ASR Confidence:
- 0.86
- Description:
-
MIT's Josh Angrist introduces differences-in-differences with one of the worst economic events in history: the Great Depression.
Economists still argue about the causes of the Great Depression, but most agree that a key piece of the puzzle was an epidemic of bank failures. Over 9,000 banks failed from 1930 to 1933!
Could the Federal Reserve have prevented this catastrophe?
At the time, regional Federal Reserve branches had considerable policy independence. Some branches helped troubled banks with “easy money”. Others did not, following a “tight money” policy.
Metrics wizards Gary Richardson and William Troost used differences-in-differences to analyze a natural experiment in Mississippi, where one half of the state had tight money while the other half had easy. What did they find?
This introduction to differences-in-differences covers the following:
- Bank failures during the Great Depression
- Easy versus tight monetary policies; moral hazard
- Parallel data trends
- Calculating the treatment effect
- Assumptions for a valid differences-in-differences analysis**INSTRUCTOR RESOURCES**
Troost/Richardson paper: https://www.journals.uchicago.edu/doi/abs/10.1086/649603
Econometrics test bank: https://mru.io/kt2
High school teacher resources: https://mru.io/o15
Professor resources: https://mru.io/t0f
EconInbox: https://mru.io/sm5**MORE LEARNING**
Try out our practice questions: https://mru.io/wfd
See the full course: https://mru.io/469
Receive updates when we release new videos: https://mru.io/7g2
More from Marginal Revolution University: https://mru.io/c30 - Video Language:
- English
- Team:
- Marginal Revolution University
- Project:
- Mastering Econometrics
- Duration:
- 11:22
Kirstin Cosper edited English subtitles for Introduction to Differences-in-Differences | ||
Kirstin Cosper edited English subtitles for Introduction to Differences-in-Differences | ||
Theresa Ranft edited English subtitles for Introduction to Differences-in-Differences | ||
Theresa Ranft edited English subtitles for Introduction to Differences-in-Differences | ||
Theresa Ranft edited English subtitles for Introduction to Differences-in-Differences | ||
Theresa Ranft edited English subtitles for Introduction to Differences-in-Differences | ||
Theresa Ranft edited English subtitles for Introduction to Differences-in-Differences | ||
Theresa Ranft edited English subtitles for Introduction to Differences-in-Differences |