Claimed’s Basement

To distract myself from admissions doomposting, I figured I’d do a bit of writing. Yes, it is supposed to read like AI wrote it.

Two predocs wake up in the basement of Littauer Center for Public Administration, chained to student desks. On each desk is a laptop sitting on top of an old copy of MWG. Everything else is grey concrete. Both predocs were members of the Opportunity Insights lab. Predoc 1 had studied economics and maths as an undergraduate at a high ranked school, but his Bs in real analysis and measure theory doomed him from going straight into a PhD. His rejection from the T10 last cycle left him with a strong inferiority complex. Predoc 2 also studied maths and economics. Coming from a lower tier state school, he never had a chance. They were captured after falling asleep at their desks. Predoc 1 having been awake for the past 48 hours cleaning data for Chetty’s latest paper. Predoc 2 after a staying up long enough to advance 10 levels in Genshin impact while waiting for his girlfriend to finish data cleaning. Although they had been working together, neither ever had the time to get to know each other.

Back in the basement, the groggy predocs rub their eyes and instinctively opened their laptops. One would assume that being chained to a chair in a basement would cause some panic, but alas, after having been in the OI lab for nearly a year, being literally chained to a desk is only a physical manifestation of their mental reality. Panic did ensue moments later. The data they’d spent the last three weeks cleaning was gone and replaced with a pdf “Predoc.Instructions” Desperate for answers, they opened the file:

Predoc,

Welcome to my basement. You have been selected to play a game for my entertainment and professional advancement. 

I wrote a paper some time ago on monetary policy and tight labour markets. Now I need you to figure out how monetary policy affects inflation different industries facing different degrees of labour market tightness.

In three minutes, you will receive an email with the paper and a wifi password. You will need to find and download 50 years of BLS inflation indices, county business patterns and QWI employment data then manually construct the longest possible panel. Every error you make will reduce your chance of admissions into a T10 programme by 2%. You have 12 hours starting now.

Completion of the task will secure your position as the top predoc, guaranteeing you the best letter of the cohort. Failure to complete the task will result in rejection from any programme, except for the MINES of course.

Claimed

3 minutes passed like 3 years. Finally, the emails arrived.

The predocs sat in their chairs, too shocked to think. Was this some kind of joke? Was this punishment? Hazing? This certainly wasn’t legal right? Or do predoc slaves not have rights? Maybe this was a dream? The latter of these hypotheses was quickly disproven by predoc 2, who tested it by dropping his copy of MWG on his foot. The only option was to play Claim’s game.

They finally came to their senses and got to work. Heads down, STATA open. Predoc 1 skimmed the paper and quickly found the data. He needed to download 50 years of files, convert county to MSA to match the data and most importantly, get the third dataset before any merging could happen. The first two steps were easy. He had done something similar for a previous project requiring him to construct a lengthy panel of health data. Within the hour, he had written a script to automate the downloading. The second set would also not be too hard, but impossible without step three. The mild mannered Predoc 1 was not one to engage in confrontation. While not particularly concerned with ethics, he was averse to anything that could jeopardise his reputation. Academic dishonesty made him sick. He did some more digging and to his surprise, did find an already cleaned version of his missing dataset on an obscure website. His delight at not needing to engage in manipulation turned to disappointment once realizing that they missed data from 1946- 1964. At 2% per month, his dreams of T10 would be gone. Anything not T10 might is the mines as far as he was concerned. Manipulating Predoc 2 for the full dataset at least could give him a shot. But then again, what if this was Chetty’s way of testing character?

Predoc 2 had access to the paper and all three datasets. Unlike Predoc 1, Predoc 2 had no choice but to manually download files one by one. It would take him well over 3 hours just to download everything. He downloaded one file and noticed something very wrong. The paper had MSA and the data was in counties. 12 hours would not be enough time to complete this, much less learn how to merge and compute summary statistics. Predoc 2’s strength was getting others to do work for him. Predoc 1 was clearly ahead. Maybe it was worth offering a trade? Claimed’s email warned against aiding Predoc 1. Predoc 2 may not have done well in game theory, but he knew enough to understand the incentive to defect, even without Claimed’s warning.

I’m going to the mines without his help, maybe he’s a good person? Might as well try. Message sent.

Predoc 1 received Predoc 2’s proposition moments later

Oh, he has no idea what he’s doing. He’ll end up in the Mines regardless of what I say. Might as well help myself

Predoc 1 felt a glimmer of hope for the first time.

Predoc 2 also got an email

Predoc 2 clicked the link. Sure enough, a low definition stream of Predoc 1’s screen showed up. Completely unreadable

FUUUUUUUUU WHYYYYYY That bastard better help me

A notification popped up

It can’t be that easy

Predoc 2 had doubts, but had no other choice. Predoc 1 got the data.

            Predoc 1

Predoc 1 ran the code. The code ran. Too well….

Wtf??

Ohhh fuqqq

Commuting zone definitions change over time. Another attempt.

Predoc 2 was getting antsy. Only 4 hours left and nothing. He could see that Predoc 1 successfully merged some of the data

*crickets*

I knew it!! GAHHH

Predoc 1 was not being as dishonest as he seemed.

SUMMARY STATS DON’T MATCH

SUMMARY STATS DON’T MATCH

SUMMARY STATS DON’T MATCH

Regardless, Predoc 1 was not going to give Predoc 2 any code. Allowing freeriding on top of ending up in the mines in too much.

It’s over…

It’s over….

2 minutes remained. Claimed decided that now was the time to increase the resolution of the screencast. Predoc 2 watched in agony from his screen as predoc 1’s final line of code matched the line on his screen. He was no closer to completing the task himself. After all, he had never done any work himself- not his schoolwork, not his application, not his work at OI. Without the full code from Predoc 1, he did not know how to disaggregate the data, much less account for changing industry codes and county borders. The jig was up. He had been played as the fool he never knew he was.

The dejected predoc 1 ran his code for a final time. He promptly closed his laptop to avoid impending doom, before he could find out that his summary statistics did indeed match that time.

Two former predocs wake up in a shared office the third floor of 816 15th street. Colorado School of Mines. Golden Colorado, 80401. They cannot leave for another 5 years.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: