2025 S2 ECOS 3997 (Stream 1: Data literacy)
Week 1: Introduction
Ellen Stuart
School of Economics
The University of Sydney
7 August 2025
Introduction
Data literacy is key for judging the reliability of research in news stories (or any
context for that matter)
1/62
Introduction
Consider these headlines that appeared in the New York Times between 2017-2019:
2/62
Introduction
Previous courses likely focused on economic theory
Economic theory is an important tool for:
• Disciplining our thinking
• Developing hypotheses about social and economic phenomena
• Helping us understand what empirical evidence is relevant
Data analysis and visualization is a process of developing empirical evidence to test
those hypotheses, distinguish competing theories, etc.
These two skills—economic theory and data analysis—are most powerful when
used together
3/62
Agenda for today
1. ECOS 3997 course logistics
• What is ECOS 3997?
• Teaching staff
• Lecture and tutorial schedule
• Assessment structure
• Some thoughts on effective communication
2. Stream 1: Data Literacy
• Topics covered and the Big Picture
• The research question and some economic theory
4/62
Course logistics: What is ECOS 3997?
ECOS 3997: Interdisciplinary Impact in Economics
• Primary goal: how do we effectively communicate economic
concepts/economic research to an interdisciplinary audience?
• There are several “streams” of ECOS 3997
• For each stream: different focus/topic, same assessment structure
5/62
Course logistics: Teaching staff
The streams taking place this semester are:
• Stream 1: Data literacy - Dr Ellen Stuart
• Stream 2: Behavioural Economics of Gambling - Dr Stephen Cheung
• Stream 3: Topics in Family Economics - Dr Chandana Maitra
• Stream 4: Implications of Macroeconomic Policies - Dr Melissa Song
I am also the course coordinator
6/62
Course logistics: Teaching staff
Instructor for Stream 1: Me (Dr Ellen Stuart—you can call me Ellen)
• Consultation hours: Fridays 9:15-10:15am, Office 560, SSB / A02
Tutor for Stream 1: Alex Berger
• Consultation hours: Mondays 10:00-11:00am, Consult Room 220, SSB / A02
7/62
Course logistics: Teaching staff
Some FAQ about me:
• Where am I from? USA
• Where in the USA? All over (but my family is from Wisconsin)
• What am I doing in Australia? Family, lifestyle, coffee
• What do I study? Behavioral reactions to taxes and tax enforcement (a subset
of Public Economics)
• Am I going to use American English? Yes (but I’m conversant in Australian
English)
8/62
Course logistics: Lectures and tutorial schedule
Lectures:
• Weeks 1-11: Thursdays 10am-12pm in SSB Lecture Theatre 200
– Lecture will start promptly at 10:05
– I will aim to end by 11:55am
– 5-10 minute break about halfway through
• Weeks 12-13: No lecture
• Phone disclaimer
Tutorials:
• Weeks 2-12 (you must attend your assigned tutorial)
• No tutorials in Weeks 1 and 13
9/62
Course logistics: Lectures and tutorial schedule
Attendance disclaimer:
• We will follow the FASS Student Attendance at Classes Provisions
• In particular, students are expected to attend a minimum of 90% of
timetabled activities for a unit of study
• Bring your Student ID to tutorials so your tutor can note you as present
10/62
Course logistics: Assessment
The assessment structure for ECOS3997 is the same for all streams:
• In-tutorial writing exercises (40%)
• Media project (30%)
• Final exam (30%)
11/62
Course logistics
If you have more questions about ECOS3997:
• First: check the ECOS3997 FAQ page on Canvas
• Second: check the ECOS3997 Ed discussion board
If your question has not been answered, ask on Ed
You should only email the teaching staff with queries that are confidential
in nature
12/62
Some comments on effective communication
This course is in part about communication
What is the most effective way to communicate with an interdisciplinary audience?
13/62
Some comments on effective communication
Stepping back: most jobs require some amount of communication
You will spend a lot of time writing in order to communicate with your boss and
colleagues
Your boss and colleagues are busy—you will be competing for their attention
How do you get people to pay attention to what you write?
14/62
Some comments on effective communication
Think about writing that has drawn you in
It probably used clear, direct language that let you be a part of the thought process
Think about writing that has felt inaccessible
It likely used lots of jargon and flowery language that you felt like you should
understand, but had a hard time following. (That’s not your fault—bad writing is
hard to follow!)
15/62
Some comments on effective communication
Some advice for writing from my friend and colleague, Ash Craig (ANU):
1. Don’t waste your first sentence: it alone should convince someone drowning in
paper to pay attention
2. Your reader should be able to know and understand the argument you are
making without having to deeply read (the “10 second rule”)
3. Be direct, not subtle or clever
4. Ruthlessly purge unnecessary words and simplify language
16/62
Some comments on effective communication
And a piece of advice from me: conveying complicated thoughts with simple
language is more impressive than conveying simple thoughts with complicated
language.
“If I’d had more time, I would’ve written you a shorter letter.” [Blaise Pascal, 17th
century, modified from the original French]
17/62
Some comments on effective communication
Dear Colleague A,
I spent the last few days implementing the list of changes that we discussed last week.
Some of the updates were a bit trickier to implement than I’d anticipated but they’re all
done now. Figures 1-7 look about like what we anticipated. They’re attached to this email.
Figure 8 on the other hand looks different. I’d like to meet to discuss this figure with
you when possible. The trends are the opposite of what makes sense given the old results
and the changes I made. I’ve gone over the code several times and looked at several case
studies and it’s not obvious to me where the error is (if there is one). You know how
helpful it is to have fresh eyes look at weird results in cases like this. Otherwise like I said
everything seems just fine.
Best,
Colleague B
Word count: 144 words
Words before key request: 58 18/62
Some comments on effective communication
Revised version:
Dear Colleague A,
Are you available to meet sometime soon?
I’ve finished the updates to the analysis that we discussed at our last meeting. The most
up-to-date figures are attached to this email.
I don’t understand the patterns in Figure 8. It would be helpful to look at the figure
together and discuss what might be going on.
The changes to the other figures are what we expected.
Best,
Colleague B
Word count: 71 words
Words before key request: 3 19/62
Agenda for today
1. Course logistics
• What is ECOS 3997?
• Teaching staff
• Lecture and tutorial schedule
• Assessment structure
• Some thoughts on effective communication
2. Stream 1: Data Literacy
• Topics covered and the Big Picture
• The research question and some economic theory
20/62
Overview of Stream 1
This stream: Data literacy
This stream will focus on (some of) the considerations of actually doing empirical
economics research
Another title for this stream could have been “how to consume research”
21/62
Overview of Stream 1
My specific goals in this stream are:
• To think about the relationship modern empirical economics research,
statistics, and data science
• To demonstrate how some of the tools and ideas you may have learned in your
other economics classes come together when undertaking empirical research
• To encourage you become more critical of any headline you read that starts
with “study shows...”
22/62
Overview of Stream 1
How will we do this?
• Part 1 (Week 1): Introduce a research question in Economics
• Part 2 (Weeks 2-6): Discuss data and data preparation
• Part 3 (Weeks 7-11): Discuss different analysis strategies
23/62
Lifecycle of empirical work
Ask a Question Obtain Data
Understand the DataPrediction and Inference
Conclusions
24/62
Research question
What is the revenue-maximizing linear income tax rate?
Before we discuss this question:
• Some motivation (why should we care about this question?)
• Some vocabulary (what does this question even mean?)
25/62
Motivation
What are some services provided by the government?
• Public education
• Healthcare
• Retirement and disability benefits
• Unemployment benefits
• Childcare subsidies
• ...
26/62
Motivation
If we believe the government should provide these services, then we should also
believe that the government should collect taxes.
Taxes provide the funds that governments use to pay for those services (a.k.a.
“transfers”)—in other words, taxes are the source of government revenue.
27/62
Motivation
To what extent should the government use taxes and transfers to reduce inequality?
Some people believe minimally/not at all, while others believe society should be
totally equal.
Of course, people also disagree about what “totally equal” means!
• Equal opportunity? Equal outcomes? (All outcomes?)
28/62
Motivation
In the classic book Equality and Efficiency: The Big Tradeoff, Arthur Okun posed
this thought experiment:
• A tax on the most affluent 5% of families finances a transfer to families in the
lowest 20% of the income distribution.
• BUT: the program has “an unsolved technological problem: the money must
be carried from the rich to the poor in a leaky bucket...so the poor will not
receive all the money that is taken from the rich.”
29/62
Motivation
What did Okun mean by “a leaky bucket”?
• There are administrative costs to collecting taxes and administering
social services
• People change their behavior in response to taxation
This second piece—the behavioral changes in response to taxation—is what leads
deadweight loss
30/62
Motivation
Deadweight loss refresher:
0
Quantity, Q
Price, P
D
S
P∗
Q∗
0
Quantity, Q
Price, P
D
S
P∗
Q∗
Consumer surplus
Producer surplus
0
Quantity, Q
Price, P
D
SS ′
0
Quantity, Q
Price, P
D
SS ′
Pc
Q∗t
0
Quantity, Q
Price, P
D
SS ′
Pc
Pp
Q∗t
t
0
Quantity, Q
Price, P
D
SS ′
Pc
Pp
Q∗t
Cons. surplus (w tax)
Prod. surplus (w tax)
0
Quantity, Q
Price, P
D
SS ′
Pc
Pp
Q∗t
Cons. surplus (w tax)
Prod. surplus (w tax)
Gov. revenue
Deadweight loss
31/62
Motivation
Your own personal tolerance for the leaks (while still supporting the program) is a
rough measure of your willingness to trade off efficiency for equity
Efficiency versus equity trade-off is at the core of Public Economics
• Efficiency: minimize the leaks (i.e., minimize deadweight loss)
−→ make the pie as big as possible
• Equity: minimize inequality
−→ make the pieces of the pie as close in size as possible
32/62
Motivation
Income inequality has been a popular topic in the news as the share of national
income held by top earners has been increasing.
4/24/2018 Australia – WID – World Inequality Database
http://wid.world/country/australia/ 1/2

Australia
POPULATION 24 158 000 | PER ADULT NATIONAL INCOME 39 614€
METHODOLOGY & LIBRARY (HTTP://WID.WORLD/METHODOLOGY/#LIBRARY-BROWSE-BY-COUNTRY-AUSTRALIA)


 FULL DATASET
Per adult national income
Per adult GDP
Per adult national wealth
Wealth-income ratio
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
KEY INDICATORS
AVERAGE INCOME
AVERAGE WEALTH
INCOME INEQUALITY
WEALTH INEQUALITY
MORE INDICATORS
Search a concept...
Per adult national income
Per adult GDP
Per adult national wealth
Wealth-income ratio
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
KEY INDICATORS
AVERAGE INCOME
AVERAGE WEALTH
INCOME INEQUALITY
WEALTH INEQUALITY
MORE INDICATORS
Search a concept...
Top 1% national income share, Australia, 1921-2014
Share (%) Share (%) More options
Sh
ar
e
of
to
ta
l (
%)
1940 1960 1980 2000
4
6
8
10
12
14
16
Pre-tax national income Top 1% share ADULTS INDIVIDUAL
Top 10% national income share, Australia, 1941-2014
Share (%) Share (%) More options
Sh
ar
e
of
to
ta
l (
%)
1950 1960 1970 1980 1990 2000 2010
25
27.5
30
32.5
35
37.5
Pre-tax national income Top 10% share ADULTS INDIVIDUAL

WORLD (HTTP://WID.WORLD/WORLD/)
BY COUNTRY
DATA (HTTP://WID.WORLD/DATA/)

METHODOLOGY ABOUT US
NEWS
EN
W O R L D
INEQUALITY
(HTTP://WID.WORLD/)
DATA B A S E
33/62
Motivation
Income inequality has been a popular topic in the news as the share of national
income held by top earners has been increasing.
4/24/2018 Australia – WID – World Inequality Database
http://wid.world/country/australia/ 1/2

Australia
POPULATION 24 158 000 | PER ADULT NATIONAL INCOME 39 614€
METHODOLOGY & LIBRARY (HTTP://WID.WORLD/METHODOLOGY/#LIBRARY-BROWSE-BY-COUNTRY-AUSTRALIA)


 FULL DATASET
Per adult national income
Per adult GDP
Per adult national wealth
Wealth-income ratio
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
KEY INDICATORS
AVERAGE INCOME
AVERAGE WEALTH
INCOME INEQUALITY
WEALTH INEQUALITY
MORE INDICATORS
Search a concept...
Per adult national income
Per adult GDP
Per adult national wealth
Wealth-income ratio
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
KEY INDICATORS
AVERAGE INCOME
AVERAGE WEALTH
INCOME INEQUALITY
WEALTH INEQUALITY
MORE INDICATORS
Search a concept...
Top 1% national income share, Australia, 1921-2014
Share (%) Share (%) More options
Sh
ar
e
of
to
ta
l (
%)
1940 1960 1980 2000
4
6
8
10
12
14
16
Pre-tax national income Top 1% share ADULTS INDIVIDUAL
Top 10% national income share, Australia, 1941-2014
Share (%) Share (%) More options
Sh
ar
e
of
to
ta
l (
%)
1950 1960 1970 1980 1990 2000 2010
25
27.5
30
32.5
35
37.5
Pre-tax national income Top 10% share ADULTS INDIVIDUAL

WORLD (HTTP://WID.WORLD/WORLD/)
BY COUNTRY
DATA (HTTP://WID.WORLD/DATA/)

METHODOLOGY ABOUT US
NEWS
EN
W O R L D
INEQUALITY
(HTTP://WID.WORLD/)
DATA B A S E
34/62
Motivation
Income inequality has been a popular topic in the news as the share of national
income held by top earners has been increasing.
4/24/2018 USA – WID – World Inequality Database
http://wid.world/country/usa/ 2/2


Per adult national income
Per adult GDP
Per adult national wealth
Wealth-income ratio
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
KEY INDICATORS
AVERAGE INCOME
AVERAGE WEALTH
INCOME INEQUALITY
WEALTH INEQUALITY
MORE INDICATORS
Search a concept...
Per adult national income
Per adult GDP
Per adult national wealth
Wealth-income ratio
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
KEY INDICATORS
AVERAGE INCOME
AVERAGE WEALTH
INCOME INEQUALITY
WEALTH INEQUALITY
MORE INDICATORS
Search a concept...
Top 1% ×scal income share, USA, 1913-2015
Share (%) Share (%) More options
Sh
ar
e
of
to
ta
l (
%)
1920 1940 1960 1980 2000
7.5
10
12.5
15
17.5
20
22.5
25
Fiscal income Top 1% share
Top 10% ×scal income share, USA, 1913-2015
Share (%) Share (%) More options
Sh
ar
e
of
to
ta
l (
%)
1920 1940 1960 1980 2000
30
35
40
45
50
55
Fiscal income Top 10% share

CONTACT US
(MAILTO:IN%66%6F@%77I%64.WOR%6C%64)
WEBSITE CREDITS FAQ
THE SOURCE FOR GLOBAL INEQUALITY DATA
DONATE

FOLLOW US SHARE THIS
BY COUNTRY
DATA (HTTP://WID.WORLD/DATA/)
METHODOLOGY ABOUT US
NEWS
35/62
Motivation
Income inequality has been a popular topic in the news as the share of national
income held by top earners has been increasing.
4/24/2018 USA – WID – W rld Inequality Database
http://wid.world/country/usa/ 2/2


Per adult national income
Per adult GDP
Per adult national wealth
Wealth-income ratio
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
KEY INDICATORS
AVERAGE INCOME
AVERAGE WEALTH
INCOME INEQUALITY
WEALTH INEQUALITY
MORE INDICATORS
Search a concept...
Per adult national income
Per adult GDP
Per adult national wealth
Wealth-income ratio
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
Top 10% share
Middle 40% share
Bottom 50% share
Top 1% share
KEY INDICATORS
AVERAGE INCOME
AVERAGE WEALTH
INCOME INEQUALITY
WEALTH INEQUALITY
MORE INDICATORS
Search a concept...
Top 1% ×scal income share, USA, 1913-2015
Share (%) Share (%) More options
Sh
ar
e
of
to
ta
l (
%)
1920 1940 1960 1980 2000
7.5
10
12.5
15
17.5
20
22.5
25
Fiscal income Top 1% share
Top 10% ×scal income share, USA, 1913-2015
Share (%) Share (%) More options
Sh
ar
e
of
to
ta
l (
%)
1920 1940 1960 1980 2000
30
35
40
45
50
55
Fiscal income Top 10% share

CONTACT US
(MAILTO:IN%66%6F@%77I%64.WOR%6C%64)
WEBSITE CREDITS FAQ
THE SOURCE FOR GLOBAL INEQUALITY DATA
DONATE

FOLLOW US SHARE THIS

36/62
Quick overview of taxation
Our research question is: What is the revenue-maximizing linear income tax rate?
Motivation: Understanding how to efficiently and equitably raise revenue to fund
government services is important and highly relevant!
Vocabulary: What is the specific type of tax our question is about?
37/62
Quick overview of taxation
There are lots of things that the government could tax to raise the revenue it needs
to provide its services
What are some examples?
• Income
• Capital (savings)
• Goods and services (consumption)
• Property / land
• Bequests and gifts
• Windows
• ...
38/62
Quick overview of taxation
39/62
Quick overview of taxation
Tax base: the activity or item on which a tax is levied (e.g., income, property,
consumption)
Our research question is: what is the revenue-maximizing linear income tax rate?
In our context, the tax base is income
40/62
Quick overview of taxation
Our research question is: what is the revenue-maximizing linear tax rate?
The “linear” part tells us the tax function we’re considering
Tax function: a function where you plug in the relevant inputs and the result is
how much tax is owed
Example: T(income, consumption, kids) = (income · 30%) + (income2 · 5%)
+ (ln(consumption) · 18%) - (kids · $1,000)
41/62
Quick overview of taxation
Average tax rate: ATR = total tax paid / total value of tax base
• Ex: if the tax base is income and the average income tax rate is 10%, then
10% of income is owed to the government
• Helps evaluate the equity of a tax system:
– Proportional tax: ATR is proportional as income increases
– Progressive tax: ATR rises with income
– Regressive tax: ATR decreases with income
42/62
Quick overview of taxation
Marginal tax rate: MTR = ∆ tax paid / ∆ value of tax base
• Ex: if MTR = 20% and individual earns an additional $1, the individual owes
an additional $0.20 to the government
• MTR impacts incentives
43/62
Quick overview of taxation
Most Western countries have a step-wise linear income tax function, where the
marginal income tax rate increases at different incomes.
For example, for the 2024-25 Australian fiscal year:
• $0 - $18,200 : Nil
• $18,201 - $45,000 : 16 cents for each $1 over $18,200
• $45,001 - $135,000 : $5,092 plus 30 cents for each $1 over $45,000
• $135,001 - $190,000 : $31,288 plus 37 cents for each $1 over $135,000
• $190,001 and over : $51,638 plus 45 cents for each $1 over $190,000
44/62
Quick overview of taxation
2024-25 Australian fiscal year:
Note, this is a progressive tax system because the ATR rises with income 45/62
Quick overview of taxation
Top marginal income tax rates have fluctuated wildly over the last century.
46/62
Quick overview of taxation
Our research question is: what is the revenue-maximizing linear tax rate?
Instead of a progressive step-wise linear income tax function, we will focus on a
simple linear income tax function (i.e., only one marginal tax rate):
T (income) = τ · income
47/62
Quick overview of taxation
How should we think about what the best τ is?
The best value for τ depends on our objective
Our research question: what is the revenue-maximizing linear income tax rate?
Our research question gives us an objective: maximize government revenue from
income taxes using a linear income tax function
48/62
Quick overview of taxation
What are some other objectives we might have as policy makers?
Economists will often use the phrase “social welfare function” (SWF)—this is a
mathematical representation of our policy objective that we as social planner are
trying to optimize
A popular social welfare function is a utilitarian SWF
• The goal for a utilitarian is to maximize the sum of everyone’s utility (i.e., the
size of the pie)
. Another social welfare function that you might hear discussed is a Rawlsian SWF
• A Rawlsian seeks to maximize the utility of the least-well-off person in society
49/62
What is the revenue-maximizing linear income tax rate?
We’ve discussed every aspect of our research question. Now let’s answer it.
Goal: Maximize revenue from a linear income tax rate
How could we model this with basic economic theory?
50/62
What is the revenue-maximizing linear income tax rate?
Set-up:
• Linear tax τ
• Individual i has pre-tax income zi = wi li
• N taxpayers, where N is large
• Average pre-tax income ≡ Z =∑i zi/N
• Per-person tax revenue ≡ R = τ · Z = τ ·∑i zi/N
– We assume this is immediately transferred back to the individuals
• Individual i consumes their post-tax income: ci = (1− τ) · zi + R
• Individual i chooses consumption ci and labor supply li to maximize utility:
ui (ci , li ) = ui
(
(1− τ) · wi li + R, li
)
51/62
What is the revenue-maximizing linear income tax rate?
Assume: people work more when the tax rate is lowered
⇒ average pre-tax income Z increases with the net-of-tax rate (1− τ)
– In other words, average pre-tax income Z is a function of the net-of-tax rate
1− τ and we write it as Z (1− τ)
⇒ per-person tax revenue R(τ) = τ · Z (1− τ) is inversely U-shaped
– As τ goes up, Z (1− τ) goes down
52/62
What is the revenue-maximizing linear income tax rate?
τ = 0⇒ R = 0
τ = 1⇒ R = 0
τ = τ ∗ −→ Rmax
53/62
What is the revenue-maximizing linear income tax rate?
Solving for τ∗:
R(τ) = τ · Z (1− τ)
R ′(τ) = Z − τ dZ
d(1− τ)
R ′(τ∗) = 0⇒
0 = Z − τ∗ dZ
d(1− τ)
Z = τ∗
dZ
d(1− τ)
1 =
τ∗
Z
dZ
d(1− τ)
54/62
What is the revenue-maximizing linear income tax rate?
Because we’re economists, we want to think about things in elasticities
Define e ≡ dZ/Zd(1−τ)/(1−τ) = dZd(1−τ) 1−τZ
What is the elasticity of taxable income, e?
The answer to: “if I lower τ so you keep one percent more of your income, by what
percentage do you increase the income you earn?”
55/62
What is the revenue-maximizing linear income tax rate?
Solving for τ∗, continued:
1 =
τ∗
Z
dZ
d(1− τ)
1 =
τ∗
Z
dZ
d(1− τ) ·
1− τ∗
1− τ∗
1 =
τ∗
Z
dZ
d(1− τ) ·
1− τ∗
1− τ∗ =
τ∗
1− τ∗ ·
dZ
d(1− τ)
1− τ∗
Z︸ ︷︷ ︸
=e
1 =
τ∗
1− τ∗ · e
τ∗ =
1
1 + e
⇒ If we know e, we know τ∗
This is something we can estimate empirically!
56/62
What is the revenue-maximizing linear income tax rate?
Some observations:
1. The revenue-maximizing linear tax rate is a function of a single empirically
estimable parameter: the elasticity of taxable income (e)
2. The revenue-maximizing linear income tax rate an upper bound on the
optimal linear tax rate
• It is Pareto inefficient to have τ > τ∗
• Decreasing τ would make taxpayers better off: they pay less taxes AND would
increase tax revenue for the government (which gets given back as R!)
57/62
What is the revenue-maximizing linear income tax rate?
Where do we go from here?
• Part 1 (Week 1): introduce a research question in Economics
We just did this—and even know a (theoretical) answer!
• Part 2 (Weeks 2-6): discuss data and data preparation
• Part 3 (Weeks 7-11): discuss different analysis strategies
For the rest of the course, we will think about how to answer our question
empirically
In other words, how could we use data to estimate e?
58/62
What is the revenue-maximizing linear income tax rate?
Part 2: focus on the initial work needed to conduct an empirical analysis
• Week 2: What is data?
• Week 3: Summary and inferential statistics
• Week 4: Data visualization
• Week 5: Data cleaning and processing
We will use our goal of estimating e as motivation for this work, but it will take a
backseat for these lectures
59/62
What is the revenue-maximizing linear income tax rate?
Part 3: different analyses strategies used in applied microeconomic research:
• Week 6: Linear regression
• Week 7: Instrumental variables (IV)
• Week 8: Difference-in-differences (DND)
• Week 9: Regression discontinuity (RD)
• Week 10: Bunching analysis
• Week 11: Catch-up
For each strategy, we’ll cover the intuition and core underlying assumptions before
discussing an academic paper that uses that technique to estimate the parameter
we need for our research question: the elasticity of taxable income
Our focus will be on how the research methodology was communicated
60/62
Wrapping up
Today: Course logistics and the research question
• Key take-away: if we can estimate e, we know τ∗
61/62
Wrapping up
Next week (Week 2):
• Lecture: What is data?
• Tutorial: (Unmarked) Prompt 0 (next slide)
References:
• UC Berkeley Data Science 100 (http://ds100.org/sp23/)
62/62
Prompt 0 (unmarked)
Think about a time when you found the results of a research study very
compelling. This could be something you read for an economics class, or in
the newspaper, or elsewhere. What made you believe the results? Was the
writing especially clear, or perhaps the visuals were impressive? What about
the description of the research methodology? Did the results align with your
priors? Similarly, think about a time when you were skeptical of the results of
a research study. Why were you skeptical?
Reminders:
• Week 2 is the best time to practice accessing Canvas via lockdown browser
• Bring your ID!

学霸联盟