For Those Headed to Atlanta Next Week..

Here is everything you need to know about attending and participating in conferences. Hilarious!!! I had never seen this handy instructional guide but I recognize the behaviors so clearly many others have studied it carefully. Luckily, many of the suggestions come naturally, and so I have in the past just naturally put many words on slides and arrived late to sessions. I'm discussing two papers at the ASSA meetings next week---I will be sure to point out all typos in the papers!  I'm sure the audience members and authors will be grateful. :)

Writing, Presenting, Discussing, Refereeing..

You know, those big parts of our jobs that nobody really teaches you. Amanda Agan gathers useful tidbits and puts them all here. I've blogged about some of these before but always useful to have a another look. UConn third year paper writers and job market candidates, this is for you!

Merry Day After Christmas!

How to Get Your 107 Word Abstract Down to 100 Words

Many journals put word limits on abstracts. Often, there are word limits on applications for grants. We should always strive not to include any extra words in our writing---readers are busy people, don't waste their time! But if the your deadline is approaching and you really need to cut a few words quickly, then read this twitter thread.

My favorite one (because I'm guilty of this):

A common writing quirk: Writing a thick, dense sentence followed by "In other words," with a lucid explanation. Get rid of the first sentence and start with the second.

Should You Control for Age or Year of Birth?

From our friends at xkcd.com.

Answer to the question: Both if you have multiple cross-sections of data.

If you have only one cross-section, then you won't be able to to distinguish the two. Be honest about that.

Best Stata Command Ever! ..at Least the Most Motivating!

You're in a slump. You have no idea how to respond to a referee. You're not sure whether you should completely abandon the project. Maybe the entire career plan. What should you do? Open up Stata (if it's not already open). Type this:

ssc install motivate
motivate

After it's installed, you can just type "motivate" into Stata on an as needed basis.

Thank you, Kabira Namit, for bringing this to the world.

How to Answer Questions During Seminar or Job Talk

This is an excellent twitter thread written by Michael Kraus, a psychologist at Yale. 

I will summarize the basic idea (and of course adding my own opinions while I'm at it). The ability to answer questions during a job talk is very important. The best way to prepare to answer questions is to..make sure you know the answers to the questions. Give as many practice job talks as you can to many different people so that the questions do not come as a surprise. Some questions come up often. Consider answering those within the talk. If that isn't possible, be sure to have a well-prepared answer to those (maybe a link in your slides to the answer). Also, know the background of  your topic. If you're studying the impact of a particular policy, know the details of that policy. Have a quick look at the most closely related papers on the night before your talk so you know the literature. 

Even if you do all of these things perfectly, there will still be questions you can't answer.  My favorite piece of advice from the thread: 

..it is okay to say you don't have an answer to a question--your data can't possibly be comprehensive enough to answer every important question. But, don't stop with "my data can't speak to that." Add what data you'd need, the analysis you might do, or the study you'd design.

Career Advice for Graduate Students and Folks on the Tenure Track

This PowerPoint was written by Amy Catalinac to show women how to overcome barriers in political science departments. The advice is excellent for women and men in economics. Many of the things she mentions I have heard before but always good to be reminded over and over again. 

My favorite advice (this one is on how to get good letters of recommendation): 

Jump at every opportunity offered, if it increases exposure to you/your work or involves an experience that advisors can write about. Comes at a bad time? The quicker you get used to that, the better! Feel you’re not capable? Then you must say yes. People who only do things they’re comfortable doing won’t reach their potential. 

I also really like the advice about framing the paper. Many graduate students believe that the hard work is doing the data analysis and preparing those tables. Yes, that is absolutely important, but it also takes A LOT of time, energy, and thought to understand the significance of the numbers in those tables. It is your job to make that significance "obvious" to the readers. The tables are not enough. 

My advice for third year paper writers at UConn: Do not wait until the week before the deadline to start writing up results. Writing is hard work. You can always do more data analysis after you have a draft of the introduction written.  My bet is that the process of writing will inspire really nice ideas for further data analysis. 

Measurement Error and Attenuation Bias

I know, I know. It's been a while since I've posted anything. Crazy-busy semester. Again. I'll be back in action soon, but in the meantime have a look at this amazing animation showing why (classical) measurement error leads to attenuation bias. It looks like you can even make the animation yourself in R. Thank you, Lionel Page, for tweeting about this and Maarten van Smeden for making it. 

It always made sense to me: Random noise has zero correlation with anything. As you add more noise to a variable, you'll get closer to that zero correlation. But to see it in an animation, that's just really cool. 

PhD Economists in Tech

Last week I had a student in my office telling me that Amazon may be his dream job. I told another student to at least look at the non-academic job postings because there can be some really interesting work outside of academia. Definitely worth taking a peek. 

It struck me, however, that I am really not sure how students should prepare for non-academic jobs in general and tech jobs in particular...well, besides writing an amazing dissertation. Good news: Susan Athey and Michael Luca just put out a new NBER working paper explaining the whole thing. My big takeaway: there is a wide variety of really interesting work done by PhD economists at tech companies. In terms of how to prepare, I will copy-paste from the article: 

"While PhD economists are well suited to tech careers in many ways, we also see areas for the field to improve the preparation of PhD economists for working with or in tech companies. First, with the importance of prediction, targeting, and precise estimates in tech companies, machine learning plays an important role in tech companies. While the field of economics has long been a leader in causal inference, the field is still in the process of incorporating machine learning into its standard toolkit. Second, economists have historically received less training, relative to computer scientists, at coding and at optimizing code to run statistical algorithms at large scale. Investing in these skills (and incorporating them into the PhD curriculum) can help to prepare economists to work in this area. At the same time, it remains important that economists have a strong conceptual understanding of economic issues like incentives and equilibrium effects, as well as strong empirical skills in the areas such as causal inference that we have described in this paper." (Athey and Luca 2018).

You might also want to have a look at this list of the 25 best companies for perks and benefits. 

Also, write an amazing job market paper. 

How Many Papers Should You Be Working On?

Answer: 6.

In case you don't believe me, here is the discussion on AEA's discussion board Econspark. My additional advice: You should probably devote more time to your most promising papers, but another thing to keep in mind is that if you've started a project that someone else is likely to also work on (because, for example, it is an obvious next step in the literature and the data is freely available), then you may want to get that paper through to the draft stage fairly quickly. 

H/T: David Mackenzie. Again. 

What Does Your Functional Form Say About You?

The folks at xkcd spell it out for us:

What Does a Difs in Difs Estimate Actually Tell Us?

Most differences in differences papers these days have a treatment that turns on and/or off at different times. A specific group can be in the treated category or the control category depending on the year. What does this imply for our interpretation of a differences in differences estimate? Does this matter? Maybe not if a particular policy has one causal impact all the time, no matter what, but if the impact of a policy changes over time (and most things in life do change), then this matters. For a detailed and formal discussion of the issues, see Andrew Goodman-Bacon's NBER working paper, "Difference-in-Differences with Variation in Treatment Timing." For an easy (and fun!) to read explanation, see his twitter thread. 

For all authors of econometrics/econometrics-y papers, I'd love to read more twitter threads like Andrew's describing the main ideas of your work. And if you can insert a gif in between, that just makes everything more enjoyable. 

Cool Data Alert: Minimum Wages

Do you need to get started on your third year paper? Are you interested in studying the impacts of minimum wage changes? Well, have I got the data set for you! This data set tracks effective minimum wage rates across U.S. states from January 1, 2011 to January 1, 2018. Perhaps even cooler, the data set also contains information on the legislative histories behind each change (for example, when it was passed by the legislature and signed by the governor). For details, see this IZA Discussion Paper. Thank you, Michael R. Strain and coauthors, for putting this together.

 

Stata Trick: Are You Doing RD?

Then download this package and you'll have everything you need to do all of the analysis!

Stata Trick: Binscatter

Sure, tables, stars, and p-values are great and, for sure, important. But when do I really believe a set of results? When am I most likely to remember results? When there is a great picture! 

Great news: The Stata command, binscatter, makes it easy to make beautiful pictures of your regression results. Thank you, Michael Stepner for sharing the code! 

Click this link to learn how to make this: 

Asked to Referee a Paper a Second Time?

Sometimes there is a very clear referee for a paper. Maybe the author of the seminal paper on a (very particular, narrowly defined) topic. Maybe one of the few researchers who have used a certain data set extensively. Maybe something else. Anyway, when this happens, the same person will often be asked to referee the same paper multiple times. Actually, now that I think of it, I've even been asked to referee papers multiple times in cases with no overly obvious link to my research. What to do when you're asked to referee a paper a second time? I have always thought that there was a clear answer to this: Just tell the journal editor you have already refereed the paper for a different journal. In my experience, editors will still want to hear your thoughts on the paper, but maybe they will try to find additional referees as well. 

I never thought this was particularly controversial, but apparently it is! See this twitter thread for the controversy and this very thoughtful discussion of the issues written by Tatyana Deryugina.

The big take-aways: 

1. If an editor (knowingly) asks you to referee a paper a second time, then the first step is to check if the authors have changed the paper. Super convenient way to do this recommended by Tatyana: draftable.com.
2. If you are an author of a paper that was just rejected, please read the referee reports carefully and consider making any changes that will improve the paper--especially the low cost changes. At the very least, please fix typos! I spend a lot of time on referee reports. When I see that authors don't bother even fixing the typos, well, it feels..insulting. I try not to take this personally and just assume that the authors are in a time crunch, but this is probably not an ideal situation for anyone. I've also refereed papers that have changed quite dramatically (for the better!) after being a rejected at a different journal. Yes, this probably makes me look favorably at that particular submission, but perhaps even more importantly, the author gains my admiration in general. 

But speaking of peer review, have a look at what the wise folks at xkcd have come up with: 

Stata 15 Cheatsheets

Sure, Stata 15 is nice but what's even more exciting? Stata 15 cheatsheets!!! Download them! Print them! Post them! Memorize them! Use them!  

For Students Going on the Market: Time for the Big Push!

Ok, maybe asking for papers drafts is a special procrastination device of mine, but it's also really important. Here is some advice on twitter for students planning on going on the market. The big takeaway: "You should have a rough draft at this point, a good sense of what to do to polish everything up, and a plan with your advisors on how to do that. If you don't - make a final push or wait until next year." 

Favorite Professor Procrastination Activity

Emailing students asking about drafts of papers at odd hours!

PhDComics comes through again.

How to Write a Paragraph

Ok, I have blogged about how to write introductions and conclusions, and also how to write the parts in between. Today: how to write a paragraph. In all honesty, I have never really thought about how to construct paragraphs. In fact, I laughed when I read LSE's Writing for Research advice because I am absolutely guilty of some of the "common paragraph problems." I often split paragraphs just because they are getting too long and struggle with what to do about one sentence paragraphs. For the official record, this often happens when referees ask me to add references, and I don't know where to put them. Take away: include all important references in the first submission. 

In any case, my main message about writing paragraphs is actually not that you need to very carefully follow any formula. What I want you to get from this post (especially my three students going on the market this coming year!) is that (a) writing is important and (b) that it is hard. I know you are busy with coding new specifications and running many robustness checks, etc., but please please, give yourself some time for writing. Job market paper writers, maybe have another look at this old post on how to write a paper. 

Some Stata Tricks: State FIPS codes and Asserting Things

Picture this: You are using individual data from one data source (actually, in my case, it was nursing home level data), but you want to merge in state-level data from a different data source. The good news: you have information on state in both data sources. The bad news: in one data set, state abbreviations are used (CT, RI..) and in the other, you have state FIPS codes. Sure, you can google "State FIPS code" and probably pretty quickly get a nice cross-walk, but there's another way. Install the ado file, "statastates" (type "ssc install statastates" into Stata). Then you just write in the type of code you have for Stata and it will add to your data set the other code types. Super easy and convenient! I used it this week. Thank you, William Schpero for sharing that ado file with the world. 

And now a secret. Or maybe a warning. Here it is: You will make coding errors. 

You can make fewer of them by being careful and checking your work, etc. But even then, you will make coding errors. Most of the time, people catch these coding errors eventually, but trust me, your life will be easier if you catch them as soon as possible after you make them. How to do that? Check your data often! I've blogged before about carefully looking at tables of descriptive statistics but here's another nifty trick: whenever possible, use the "assert" command in Stata. Nick Eubank provides a nice description of how to use this effectively. 

PSA: Read Stata Help Files Carefully

I would say that the collapse command in Stata is surely one of the commands I use most in my work. This week I remembered one of the classic mistakes I've made within this command, so I thought I'd give all of you Stata beginners a nice little warning. 

What I wanted to do? Create a new variable showing the number of non-missing observations within particular cells (let's say, country of origin-state-year cells). One way to do this: 

bysort country state year: generate N_countrystateyear=_N 

but also easy enough to create this variable within a collapse command. Now, you would be tempted to use the 'count' option since the help file says that this will "count the number of nonmissing observations."  

generate x=1 

collapse (count) N_countrystateyear=x=income, by(country state year) 

This will work perfectly if you are not using weights. However, if you are using weights (pweights in particular), the count option will instead give you the sum of the weights over observations in the group.  Yes, this is explained in the help file but all the way at the bottom. Hence, my warning to read those things carefully. Also, always just look at the variables you create to make sure they make sense. So if you have weights and need them to create summary statistics of your other variables, what to do? Use the 'rawsum' option. For example, 

generate x=1 

collapse (mean) age schooling etc (rawsum) N_countrystateyear=x  [pweight=perwt], by(country state year) 

For more information on the different types of weights you can use in Stata, see here. Short version: Use pweights. Sometimes, Stata will refuse to calculate something using pweights. Ever wonder why? See here. 

PS

Thank you Kerry Papps and Nikos Theodoropoulos for encouraging me to write this blog post!  

Tips 4 Economists

Masayuki Kudamatsu has put together a gigantic list of bits of advice for economists. The best part: it's not just for graduate students and folks on the tenure track. It contains advice for people becoming department heads, editors of journals, and even deans! 

Are You Writing a Difs in Difs Paper Exploiting Policy Variation Across U.S. States?

It doesn't matter which policy you're evaluating or what outcome you're considering, referees and seminar participants will always be concerned that states adopting a new policy will by coincidence (or not so much by coincidence) adopt other policies at the same time that may be driving your results. Another possibility is that states adopt new polices in response to changing characteristics of the population---and these changes are driving the variation of your outcomes. It is impossible for researchers to control for all changes in policies and all demographic characteristics, but we can certainly assuage concerns by adding to our models controls for changes in other policies or demographic characteristics. And good news: thanks to the folks at IPPSR, you can find data on many of these things in one easy to use data set! Below is the suggested citation and you can click on the link for more information: 

Suggested Citation:

Jordan, Marty P. and Matt Grossmann. 2017. The Correlates of State Policy Project v.2.0. East Lansing, MI: Institute for Public Policy and Social Research (IPPSR). 

PS

Advanced undergrads and beginner graduate students looking for data and a paper topic for a term paper, I bet that perusing through the data may inspire many great ideas! 

And What About After Tenure?

Yes, this blog is written mostly for graduate students and young assistant professors. This is because I have more advice to give on those matters. But life (and work and goals and dreams) does not end when you get tenure, even if it's at a university you really like. There is still more work to do. The only difference is that people pay less attention to you. People aren't as concerned about protecting you from difficult service assignments. No more course releases. It's not that these things are not available, it's just that now you're responsible for getting them for yourself. And by now, you should be able to. That said, my firm belief is that we're never too old to get some good mentoring. 

This morning, Susan Dynarski tweeted about when associates should go up for full. Thank you for that! Associates, see the full thread here. Full professors, feel free to provide more advice in the comments below...or email/call/message/stop by my office/etc. to share your thoughts with me personally. 

Happy Memorial Day Weekend!

From the folks at phdcomics.com.

Why We Have Deadlines...

Two of my students are defending their dissertation proposals tomorrow. It seems kind of funny to be defending proposals in front of a committee of three plus two outside reviewers, but I'm very pleased at how much progress they've made while preparing for this. 

And now I need to find a good deadline for myself.

Summertime = Research Time

I thought now would be a good time to share some advice on publishing shared by Jessica Hoel after she participated in the 2017 CeMent Mentoring workshop. Ladies, you can apply for the workshop here. I participated in 2010 and have only good things to say. 

Regardless of your gender or stage of career, I think you can benefit from reading these pieces of advice on publishing. Thank you, Jessica, for sharing!  

Some of my favorites: 

1. Do not work on teaching during any sort of leave (summer, parental, sabbatical).
2. Be realistic about the kind of work you can do. Design a research agenda that fits your institution and your life.
3. Think BIG. You can do little things that show how clever you are. Or you can do things that matter. What will you do with the answer? Work on things for which the answer matters.
4. You can’t get published if you don’t submit. Mentors suspect that the biggest difference between successful and unsuccessful people is how often they submit and where.
5. Abstract, Intro, and Tables do need to be flawless. It will earn you the benefit of the doubt because you’ve signaled you are a careful researcher.

And now a challenge for you: Adapt the words to this favorite summertime song of mine for an academic's summer. 

Wishing all of you a productive and/or (but hopefully and) fun summer! 

Do You (or Someone You Know) Use a Shift-Share Instrument?

I know it's been a while, but I'm back! Grades submitted. Late referee report submitted. And I even submitted a new paper to a journal today. I'm ready to get back to blogging! 

What better way to celebrate my return than to discuss two new papers (Goldsmith-Pinkham et al. 2018 and Jaeger et al. 2018) on using shift-share (AKA Bartik, AKA Card 2001) instruments! I have read one of these rather carefully but haven't yet gotten to the other. In any case, David Mackenzie has come through once again by providing a nice intuitive explanation of both papers. Read the blog entry but also read Tim Bartik's very careful response to the blog entry and the two papers. Yes, the very same Bartik of the Bartik instrument. 

My remarks: If we were going to limit ourselves to only writing papers with absolutely no shred of concern about identification, I think we would all be running RCTs. That might be fun but it would substantially limit the number of important questions we could answer (albeit imperfectly). As Tim Bartik writes, the question is often not whether the IV is flawless but whether it is better than the OLS. 

All of that said, I absolutely agree and feel very strongly that authors have a responsibility to know the limitations of their work and to be very straightforward about those limitations. Goldsmith-Pinkham et al. 2018 and Jaeger et al. 2018 have each provided some important tools for evaluating our IV analyses. As a frequent user of these instruments, I say thank you. But please don't start rejecting my papers. ;)  

It's That Time of Year...

I know, I know. I haven't been blogging. I haven't been responding to emails as fast as I usually do. It's just a crazy-busy time of year. Aaak, I just remembered I need to do my taxes this weekend! The good news is that the end of the semester is in sight. Until then, I'll leave you with this comic from phdcomics.com:

Writing Productivity Boosters

This LSE Impact blog post has some nice tips for increasing writing productivity. I think they work just as well for writing Stata dofiles as they do for writing papers. My favorite suggestion is the first one. I tend to do all of things that end up on my Outlook calendar (meetings with students, dentist appointments, etc.). Why not block out time for research as well?

Fake LaTex

I really hate to be a proponent of fake anything. But it's so much easier to offer comments on students' papers written in word, and yes, I know that some people take LaTex people more seriously than Word people. So, here it is: Instructions on how to make your Word document not look like a Word document. Use it if you must. But for the official record, I'm very happy with Word documents that look like Word documents. H/T: David McKenzie and Marc Bellemare. 

More on Synthetic Control Methods

Arindrajit Dube has an excellent twitter thread on synthetic control methods. Super-informative throughout. I really like the discussion at the end with Wojtek Kopczuk who writes, "My preference, when possible, is to first see a variant of diff-in-diff (as simple as practical for the context) to show the presence of an effect and only then run SC to get precision. Then you hopefully get both transparent identification and tight estimates." 

PDF Document to Excel Sheets, No Problem

Let's say you find the exact data you need, but it's in a lovely PDF report.  Do you hunker down and copy the numbers into an excel sheet--possibly making mistakes? Win a grant so you can pay someone else to do it--who again may make mistakes? No need. There's an ap for that! See here. If the link ever stops working, google "PDF Table Extractor".  

Why I Never Blog About Theory

Francesco Amodio just tweeted the answer:

PowerPoint to Beamer, No Problem

Job market candidates use Beamer. It definitely looks nice. Also, as part of the audience, I really like to know how many slides and concepts remain. If there is a lot left, I leave my slightly irrelevant questions for after the talk. That said, PowerPoint is so easy! Especially when going back and forth with coauthors making changes. 

Solution: Jason Kirwin just blogged about a nice, easy way to convert PowerPoint slides to Beamer. Just click here. 

How to Avoid Making Stata Coding Errors

OK, I'm not sure whether they can be avoided completely, but these are definitely good tricks for minimizing how often they occur and spotting them quickly. Thank you, Tal Gross, for putting them together. 

My favorites: Use sensible names for variables and programs. Comment everything! Always keep the original data set untouched in a nice safe place. 

Source: phdcomics.com

Best Stata Tricks Twitter Thread Ever

Click here for the thread. Thank you, Seth Gershenson, for putting it together. Thank you, David Mckenzie, for bringing it to my attention. Grad students, study these carefully. Folks who have been coding Stata for years, you can also find tricks to simplify your life considerably!

My favorites:

If you're plotting regression coefficient estimates & confidence intervals this fine saturday morning, coefplot is the command for you.

binscatter by @michaelstepner: plots mean of y for different bins of x. super helpful for visualizing correlations, esp in big sample when plotting raw data yields a blotch of ink. also for plotting RD reduced forms & event studies. 


Enjoy!

How to Write a Paper: The Part in Between the Introduction and Conclusion

I have blogged before about Keith Head's formula for writing an introduction as well as Marc Bellemare's formula for writing a conclusion. Marc has come through for us once again with a formula for writing all of the bits in between. Like Marc, I always start my papers with section titles. (Hmm..maybe it makes me feel like the paper is practically written, I just need to fill in some words.) Next, I take a stab at the introduction. This helps guide my research, think about a structure, etc.. After that, I start writing up different sections often based on my mood. But here's the key thing: after writing each section, I typically go back to the introduction. The introduction gets written and rewritten many, many times through the course of writing a paper. Yes, introductions are THAT important. The good news: no need to worry about getting it perfect the first time. 

Back to the structure of the paper. What to do when in the same paper you use different data sets and/or different empirical frameworks? Sometimes authors describe both data sets in the data section, then both estimating equations in the empirical framework section, etc. The problem with this approach is that by the time results are presented, readers have forgotten some of the details about the data or estimating equation. I prefer to start the paper with baseline results, then I describe any additional data sets and estimating equations (and further results) after that. This certainly makes for less boring repetition. 

How to Publish in AEJ: Applied

I also considered titling this post: what makes a great paper (but I had already used that title). 

David Deming is starting as co-editor of AEJ: Applied and has recently tweeted his thoughts on how he thinks about where a paper should be published. This week, I have spent a good chunk of my time advising students on which of their paper ideas are most promising to pursue as dissertation topics. David's tweets provide excellent guidance on this. The main thing:  

"The primary question I have in mind when reading a paper is “what have I learned from reading this that I did not already know?” This is a heuristic that helps me weigh impact/contribution vs. methodological strength. For RCTs and quasi-experiments - if the methodological approach is not convincing, it is hard to feel I’ve learned anything, so the question of impact is moot. However, if the question is uninteresting or has been studied many times before, even airtight identification still leaves me feeling like I have not learned anything So methodological strength and importance of question are strong complements." 

Read the entire thread! He also discusses his own very positive experience with a journal editor, and hopes to do the same. For all of you journal editors who help authors transform their "raw" ideas into excellent papers, thank you. On behalf of the entire profession. 

Running of the Economists

To the graduate students who don't yet know about what it takes to get your first job, watch this. To those of you who have already gone through it, be prepared for it to bring back some pretty traumatizing memories. To those of you who are going through it right now, DURING A BOMB CYCLONE, good luck!!!!