This paper/thread has the potential to provoke incorrect quick reactions that “market access regressions are now wrong.” It proposes an approach, useful if its assumptions are plausible in an empirical context, but if not then use the standard approach (and its assumptions). 1/9
This paper assumes some railroad lines are built randomly (instead of known counterfactual lines). Then, it would be natural to compare places whose market access increases more from randomly built RRs than from counterfactual RRs. Useful wherever that setup exists. 2/9
Jun 20, 2020 • 4 tweets • 2 min read
Grad students with #econhistory ideas: I’ll send you feedback if you email me 1 page about the question, data (available or acquirable), research design, main concerns, and how you aim to address those. Multiple 1-pagers better. All encouraged, please spread word. #EconTwitter
Clarifications: the goal is finding ideas that "could" become your PhD job market paper and be published in “general interest” econ journals – you can do it! You may be pre-PhD... but that’s the scale/scope expected. Junior faculty can also reach out. Question-Data-Approach.
