3rd Year Workshop

class: center, middle, inverse, title-slide

.title[
# 3rd Year Workshop
]
.subtitle[
## Write better
]
.author[
### Ian McCarthy, Emory University
]
.date[
### Economics PhD Professionalism Workshop
]

---

<style type="text/css">
.remark-slide-content {
    font-size: 30px;
    padding: 1em 2em 1em 2em;    
}
.remark-code, .remark-inline-code { 
    font-size: 20px;
}
</style>

# How to write?

I wish I knew!

--
But there are some basic rules that might help.

---
# Introduction

The introduction is critical. It is the most important thing you'll write for any paper. Make it awesome!

---
count: false

# Introduction [formula](https://www.albany.edu/spatial/training/5-The%20Introduction%20Formula.pdf)

1. **The Hook!** Why should anyone care about this at all? What big issue are you dealing with?
2. **Question.** What are you actually going to do? What is your actual research question? Doesn't need context.
3. **Context.** What are the key things that other people have done? What literature should a reader know in order to understand your contribution? 
4. **Contribution.** What are you doing differently? Should be in context of the existing literature.
5. **Expand.** I've added this. Modern applied papers (for better or worse) need to talk about some key additional findings in the intro (mechanisms, robustness, and policy implications, etc.).

---
# Paper structure

We've all seen the "standard" applied econ paper structure:

.pull-left[
1. Introduction
2. Theoretical motivation or institutional details
3. Data
4. Estimation
5. Results
6. Conclusion
]

---
count: false

# Paper structure

We've all seen the "standard" applied econ paper structure:

.pull-left[
1. Introduction
2. Theoretical motivation or institutional details
3. Data
4. Estimation
5. Results
6. Conclusion
]

.pull-right[
I **do not** like this formula...
- Fine as outline
- Too rigid and boring
- Need better **story**
]

---
count: false

# Paper structure

- Take a look at (almost) any top applied paper in the last several years
- Most will start similarly (intro, theory/policy details, maybe data)
- Deviate based on relevant analyses and estimation strategies

---
# Writing steps

Following Jesse Shapiro's [four steps to an applied paper](https://www.brown.edu/Research/Shapiro/pdfs/foursteps.pdf)...

1. **Aspirational** introduction (we've already done this)
2. Research (not part of this workshop)
3. Be a **robot**

--
This means you are **linear**, **clear**, **plain**, and **formal**. Return to (2) as you identify gaps.

---
count: false

# Writing steps

Following Jesse Shapiro's [four steps to an applied paper](https://www.brown.edu/Research/Shapiro/pdfs/foursteps.pdf)...

1. **Aspirational** introduction (we've already done this)
2. Research (not part of this workshop)
3. Be a **robot**
4. **Contractual** introduction

--
Now re-write the intro with your new knowledge. This should be a contract that spells out what you will and won't do in the paper. Return to (2) and (3) as needed.

---
# Storytelling

- Like it or not, you have to **sell** your work
- Can be hard and can feel antithetical to the "objective" scientific process
- But we're humans and we need to be entertained!

.center[
  ![](https://media.giphy.com/media/hrnYspWWhsIyA/source.gif)
]

---
count: false

# Storytelling

So how do we write something compelling? Try [paper writing gone hollywood](https://science.sciencemag.org/content/355/6320/102)...

- A general outline isn't enough
- Need to "storyboard" each section or subsection
- Rely heavily on outlines rather than getting bogged down in verbiage

---
# Some rules

Finally, let's close things out with some simple [rules](http://home.bi.no/charlotte.ostergaard/students/tenruleswriting.pdf), courtesy of Claudia Goldin and Lawrence Katz.

---
count: false

# Some rules

I'll add a few to these...

1. Be careful with self-plagiarism (also with normal plagiarism)
2. Write the damn abstract! **Do not** copy and paste from the intro, or vice versa.
3. Use "I" if you're the only author. Use "we" otherwise.
4. Define the acronym the first time only, then just use it.

--
Seriously, don't copy and paste your own words from another paper. Be more creative than that. And definitely don't copy someone else's words.

---
# Some examples

Below are a few links to papers that are generally thought to be extremely well-written:
- [Chay and Greenstone (2005), JPE](https://www.jstor.org/stable/10.1086/427462)
- [Gentzkow and Shapiro (2008), QJE](https://web.stanford.edu/~gentzkow/research/TV_QJE_2008.pdf)
- [Einav, Finkelstein, and Cullen (2010), QJE](https://academic.oup.com/qje/article/125/3/877/1903679)
- [Weitzman (2009), ReStat](https://scholar.harvard.edu/weitzman/publications/modeling-and-interpreting-economics-catastrophic-climate-change)