Return to Video

Introduction to Differences-in-Differences

  • 0:00 - 0:03
    - [Teacher] The path
    from cause to effect
  • 0:03 - 0:06
    is dark and dangerous,
  • 0:06 - 0:09
    but the weapons
    of econometrics are strong.
  • 0:09 - 0:14
    Wield differences-in-differences
    when witnessing parallel trends.
  • 0:14 - 0:17
    ♪ [music] ♪
  • 0:20 - 0:21
    Masters of metrics
  • 0:21 - 0:25
    look for convincing
    ceteris paribus comparisons.
  • 0:25 - 0:29
    The ideal comparison contrasts
    treatment and control groups
  • 0:29 - 0:31
    that look similar.
  • 0:31 - 0:35
    But sometimes this sort
    of comparability is elusive.
  • 0:35 - 0:37
    When treatment and control groups
  • 0:37 - 0:40
    evolve similarly
    in the absence of treatment,
  • 0:40 - 0:42
    even if from different
    starting points,
  • 0:42 - 0:45
    there's hope for causal inference.
  • 0:46 - 0:48
    The weapon that exploits
    parallel evolution,
  • 0:49 - 0:51
    masters say "parallel trends,"
  • 0:51 - 0:53
    is called differences-in-differences...
  • 0:53 - 0:54
    - [Man whispering]
    Differences-in-differences...
  • 0:54 - 0:57
    - ...or DD for short.
    - [Man] Alright.
  • 0:57 - 1:00
    - Let's see how DD
    can help us understand
  • 1:00 - 1:03
    one of the most important
    economic events
  • 1:03 - 1:04
    in U.S. history.
  • 1:05 - 1:08
    - [Joshua] Look back with me now
    at the Great Depression --
  • 1:09 - 1:12
    the worst economic catastrophe
    our country has ever known.
  • 1:13 - 1:16
    Unemployment hit 25% in 1933 --
  • 1:17 - 1:19
    a level not seen before or since.
  • 1:19 - 1:22
    Millions lost their homes
    or their land.
  • 1:23 - 1:25
    Suicide spiked, and hungry families
  • 1:25 - 1:27
    relied on soup kitchens
    and bread lines
  • 1:27 - 1:28
    to keep from starving.
  • 1:29 - 1:32
    - Economists argue fiercely
  • 1:32 - 1:34
    over the causes
    of the Great Depression.
  • 1:34 - 1:37
    Most agree, however,
    that a key piece of the puzzle
  • 1:37 - 1:39
    is an epidemic of bank failures.
  • 1:40 - 1:42
    This was before deposit insurance.
  • 1:42 - 1:47
    So if your bank went bankrupt,
    your savings disappeared with it.
  • 1:47 - 1:48
    - [Cashier] Closing your account?
  • 1:48 - 1:49
    - [Customer] Yes, sir.
    I'm closing my account.
  • 1:49 - 1:51
    I wouldn't leave a nickel
    in this bank.
  • 1:53 - 1:56
    - Faced with a banking crisis,
    the Central Bank has a choice:
  • 1:56 - 1:59
    lend freely to troubled banks
  • 1:59 - 2:01
    or stand aside and refuse to lend.
  • 2:02 - 2:05
    Lending freely to banks in trouble
    is called "easy money."
  • 2:05 - 2:08
    Refusing to lend
    is called "tight money."
  • 2:10 - 2:13
    - [Joshua] Monetarist masters
    Milton Friedman and Anna Schwartz
  • 2:13 - 2:15
    famously called
    the Great Depression
  • 2:15 - 2:16
    the "Great Contraction,"
  • 2:17 - 2:18
    accusing the Federal Reserve
  • 2:18 - 2:21
    of inflicting a misguided policy
    of tight money
  • 2:21 - 2:24
    on the nation's teetering
    financial institutions.
  • 2:24 - 2:26
    They argued that easy money
  • 2:26 - 2:28
    would have kept
    many banks in business,
  • 2:28 - 2:30
    shortening the Great Depression.
  • 2:30 - 2:32
    But others disagree!
  • 2:32 - 2:34
    If banks are insolvent
  • 2:34 - 2:36
    because of unwise
    lending decisions,
  • 2:36 - 2:39
    then bailouts just encourage
    more foolishness.
  • 2:40 - 2:43
    Economists called this problem
    "moral hazard."
  • 2:43 - 2:46
    The debate over bailouts
    and moral hazard continues today.
  • 2:46 - 2:49
    Should financial behemoth
    Lehman Brothers
  • 2:49 - 2:52
    had been allowed to fail
    on the eve of the Great Recession,
  • 2:52 - 2:55
    in an ideal world,
    we'd answer this question
  • 2:55 - 2:58
    by applying different Fed policies
    to randomly selected regions.
  • 2:59 - 3:00
    But we can still learn a lot
  • 3:00 - 3:02
    by using differences-in-differences
  • 3:02 - 3:06
    to compare trends across areas
    with different monetary policies.
  • 3:11 - 3:13
    - [Camilla] How's that even possible?
  • 3:13 - 3:16
    Don't the same Fed policies
    apply to all banks in the U.S.?
  • 3:16 - 3:17
    - [Man] Yeah.
    - Good question.
  • 3:18 - 3:21
    The Federal Reserve System
    is divided into 12 districts,
  • 3:21 - 3:24
    each headed by a regional bank.
  • 3:24 - 3:27
    Today, Fed policy is set
    at the national level.
  • 3:27 - 3:32
    But in the 1930s, regional Feds
    could do pretty much as they liked.
  • 3:32 - 3:33
    - [Man] Ah, interesting.
  • 3:33 - 3:36
    - And here's what's
    so awesome about that.
  • 3:36 - 3:39
    In 1930, the Atlanta Fed,
    running the 6th district,
  • 3:39 - 3:41
    followed an easy money policy,
  • 3:41 - 3:45
    sending wheelbarrows of cash
    to rescue insolvent institutions.
  • 3:46 - 3:49
    The St. Louis Fed,
    running the 8th district,
  • 3:49 - 3:51
    followed a tight money policy.
  • 3:51 - 3:54
    "Let fail the foolish!"
    they said in St. Louis.
  • 3:54 - 3:58
    And so a natural experiment
    in monetary policy was born.
  • 3:59 - 4:02
    Even better, this is
    a within-state experiment.
  • 4:02 - 4:04
    The border between the 6th
    and the 8th districts
  • 4:04 - 4:07
    ran smack through
    the middle of Mississippi.
  • 4:07 - 4:09
    So northern Mississippi
    had tight money,
  • 4:09 - 4:12
    while southern Mississippi
    had easy money,
  • 4:12 - 4:15
    but under the same state laws
    and banking regulations in both.
  • 4:16 - 4:17
    - [Teacher] The treatment group
  • 4:17 - 4:20
    is the district 6 part
    of Mississippi,
  • 4:20 - 4:23
    which had access to easy money
    during the crisis.
  • 4:24 - 4:25
    The control group
  • 4:25 - 4:28
    is the district 8 part
    of Mississippi,
  • 4:28 - 4:30
    which had tight money
    during the crisis.
  • 4:31 - 4:34
    The key year
    in our natural experiment
  • 4:34 - 4:35
    was 1930.
  • 4:36 - 4:37
    Caldwell & Company,
  • 4:37 - 4:40
    a massive financial empire
    in the South,
  • 4:40 - 4:42
    came crashing down.
  • 4:43 - 4:46
    Banking is a business
    built on confidence and trust.
  • 4:46 - 4:49
    The Caldwell meltdown
    caused a panic
  • 4:49 - 4:53
    that led to a widespread
    bank run all at once.
  • 4:53 - 4:55
    Depositors wanted their money back,
  • 4:55 - 4:58
    causing banks to go bankrupt
    and shut their doors.
  • 5:01 - 5:03
    We'll use differences-in-differences
  • 5:03 - 5:07
    to measure the effect
    of contrasting monetary policies
  • 5:07 - 5:09
    in response to the Caldwell crisis.
  • 5:12 - 5:16
    This figure plots the number
    of banks in Mississippi by year,
  • 5:16 - 5:19
    for the 8th and 6th districts.
  • 5:19 - 5:21
    Let's start in 1929,
  • 5:21 - 5:24
    a year before the Caldwell crash.
  • 5:24 - 5:28
    There are 169 banks
    open in the 8th,
  • 5:28 - 5:31
    and 141 banks open in the 6th.
  • 5:31 - 5:33
    Over the next year,
  • 5:33 - 5:37
    we see a similar handful
    of banks fail, in both districts.
  • 5:37 - 5:40
    The change in the number
    of banks in operation
  • 5:40 - 5:42
    is remarkably similar --
  • 5:42 - 5:45
    that's what parallel trends look like.
  • 5:46 - 5:49
    In November 1930, Caldwell crashes,
  • 5:49 - 5:50
    and the panic begins.
  • 5:51 - 5:54
    Banks failed frequently
    in the 8th district,
  • 5:54 - 5:55
    which had tight money.
  • 5:56 - 5:59
    But the decline is slower
    in the 6th district,
  • 5:59 - 6:00
    which had easy money.
  • 6:01 - 6:03
    Diverging trends in this period
  • 6:03 - 6:07
    might be attributable
    to easy versus tight money.
  • 6:07 - 6:12
    In July of 1931, the 8th district
    abandons tight money,
  • 6:12 - 6:14
    so now both districts are easy.
  • 6:15 - 6:17
    Parallel trends are restored.
  • 6:17 - 6:19
    In a counterfactual world,
  • 6:19 - 6:22
    where the 6th district
    follows a tight money policy,
  • 6:22 - 6:24
    what might have happened?
  • 6:24 - 6:29
    If we extrapolate the trend
    of the 8th district to the 6th,
  • 6:29 - 6:30
    it would look like this.
  • 6:30 - 6:33
    So the treatment-effective
    easy money
  • 6:33 - 6:36
    is how much the 6th district
    deviated from the path
  • 6:36 - 6:39
    implied by the 8th district trend.
  • 6:41 - 6:44
    How many banks
    did the easy money treatment save?
  • 6:44 - 6:49
    This table reports data
    for the treatment group, district 6,
  • 6:49 - 6:50
    in the first row,
  • 6:50 - 6:54
    and data for the control group,
    district 8, in the second row.
  • 6:54 - 6:57
    The first column shows
    the number of banks in business
  • 6:57 - 7:01
    before the crisis began in 1930.
  • 7:01 - 7:04
    The second column shows 1931.
  • 7:04 - 7:06
    This is the key period
  • 7:06 - 7:09
    when each district
    had differing monetary policies
  • 7:09 - 7:10
    during the crisis.
  • 7:10 - 7:14
    The rightmost column
    reports changes within the district.
  • 7:14 - 7:20
    District 6 lost 14 banks,
    while district 8 lost 33.
  • 7:20 - 7:24
    The mathematical formula
    for the treatment effect is simple.
  • 7:24 - 7:28
    We subtract the change in banks
    in operation in the 8th district
  • 7:29 - 7:32
    from the change in banks
    in operation in the 6th.
  • 7:33 - 7:36
    Hence, the name
    "differences-in-differences."
  • 7:37 - 7:42
    -14 minus -33 equals 19.\]
  • 7:42 - 7:47
    We estimate that 19 banks
    were saved by easy money.
  • 7:47 - 7:51
    In practice, tables and figures
    like those shown here
  • 7:51 - 7:53
    are the beginning
    rather than the end
  • 7:53 - 7:54
    of a DD analysis.
  • 7:55 - 7:57
    The problem of how to gauge
  • 7:57 - 8:00
    the statistical significance
    of DD estimates
  • 8:00 - 8:02
    turns out to be exceedingly tricky,
  • 8:02 - 8:06
    and a regression is typically
    part of the solution.
  • 8:09 - 8:12
    The key assumption
    behind a valid DD analysis
  • 8:12 - 8:15
    is that of parallel trends.
  • 8:15 - 8:18
    Recall the principle
    of ceteris paribus.
  • 8:18 - 8:21
    Our ideal comparison
    would have the two districts
  • 8:21 - 8:24
    experience an identical
    business environment,
  • 8:24 - 8:26
    except for one factor:
  • 8:26 - 8:28
    easy or tight money.
  • 8:29 - 8:32
    Both districts would have
    identical types of customers
  • 8:32 - 8:35
    who would go bankrupt
    at exactly the same rate.
  • 8:36 - 8:39
    The skill of their employees
    would be equal, and so on.
  • 8:39 - 8:43
    Perfect ceteris paribus comparisons
    would allow us to clearly see
  • 8:43 - 8:47
    the causal effect
    of different Fed policies.
  • 8:47 - 8:49
    In this case, that's not possible.
  • 8:49 - 8:54
    But the idea of parallel trends
    is based on a similar concept.
  • 8:54 - 8:57
    If we see that the two regions
    experience similar trends
  • 8:57 - 9:00
    in the number of banks over time,
  • 9:00 - 9:01
    in the absence of treatment,
  • 9:01 - 9:04
    we can assume
    they are good comparisons.
  • 9:04 - 9:07
    We see that the two districts
    move in parallel,
  • 9:07 - 9:10
    both before the crisis and after,
  • 9:10 - 9:12
    when they have the same Fed policy.
  • 9:13 - 9:16
    The only time the districts
    behave differently
  • 9:16 - 9:18
    is when the Fed policy is different.
  • 9:19 - 9:21
    In view of this,
  • 9:21 - 9:24
    Fed policy is a likely cause
    of diverging trends
  • 9:24 - 9:27
    from 1930 to 1931.
  • 9:28 - 9:30
    But we should also check
    for other changes
  • 9:30 - 9:32
    unique to northern Mississippi.
  • 9:32 - 9:33
    - [Man] Huh?
    - What do you mean?
  • 9:34 - 9:35
    - [Teacher] Imagine that bad tornadoes
  • 9:35 - 9:39
    hit northern but not
    southern Mississippi in 1930.
  • 9:40 - 9:42
    These tornadoes devastate farms,
  • 9:42 - 9:44
    causing farmers
    to default on loans,
  • 9:45 - 9:47
    which drives their banks
    out of business.
  • 9:47 - 9:49
    Then the 6th and 8th districts
  • 9:49 - 9:52
    would differ in not one
    but two ways:
  • 9:53 - 9:55
    Fed policy and weather.
  • 9:55 - 9:58
    And we'd have trouble
    identifying Fed policy
  • 9:58 - 10:02
    as the causal factor
    behind increased bank failures
  • 10:02 - 10:03
    in the 8th.
  • 10:03 - 10:04
    - [Man] Ceteris is not paribus.
  • 10:07 - 10:09
    - DD credibility lives or dies
  • 10:09 - 10:11
    with the claim that the only reason
  • 10:11 - 10:14
    northern Mississippi
    was special in 1930
  • 10:14 - 10:16
    is differing regional Fed policy.
  • 10:17 - 10:21
    We're in DD heaven with strong,
    visual evidence of parallel trend.
  • 10:21 - 10:26
    - In general, the first step
    in evaluating whether to use DD
  • 10:26 - 10:30
    is usually this type of visual
    confirmation of parallel trends
  • 10:30 - 10:32
    outside of the period,
  • 10:32 - 10:35
    when we expect to see
    a treatment effect.
  • 10:35 - 10:37
    The treatment in our example
  • 10:37 - 10:40
    is easy money
    in the face of bank failures.
  • 10:40 - 10:45
    Metrics masters use DD
    to explore effects of many policies,
  • 10:46 - 10:48
    like the minimum legal drinking age,
  • 10:48 - 10:52
    and environmental changes,
    like access to clean water.
  • 10:53 - 10:54
    In our next video,
  • 10:54 - 10:57
    we'll see an example
    of how regression is used
  • 10:57 - 10:59
    to implement a DD approach.
  • 11:01 - 11:02
    - [Narrator] Are you a teacher?
  • 11:02 - 11:06
    Click to explore ways
    to use these videos in class.
  • 11:06 - 11:09
    If you're a learner,
    make sure this video sticks
  • 11:09 - 11:11
    by taking a few quick
    practice questions.
  • 11:12 - 11:14
    Or if you're ready,
    click for the next video.
  • 11:15 - 11:17
    You can also check out
    MRU's website
  • 11:17 - 11:20
    for more courses,
    teacher resources, and more.
  • 11:20 - 11:22
    ♪ [music] ♪
Tytuł:
Introduction to Differences-in-Differences
ASR Confidence:
0.86
Opis:

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

more » « less
Video Language:
English
Team:
Marginal Revolution University
Projekt:
Mastering Econometrics
Duration:
11:22

English subtitles

Revisions Compare revisions