5 ms·

The very title of this paper suggests a deeply flawed conclusion. A sample size of one pandemic, from 100 years ago even, is used to generalize a conclusion tha

The very title of this paper suggests a deeply flawed conclusion. A sample size of one pandemic, from 100 years ago even, is used to generalize a conclusion that "Pandemics Depress the Economy, Public Health Interventions Do Not". Sure, there were multiple cities with varying levels of interventions, but you can only conclude that in the 1918 economy for that specific pandemic, the health interventions were good for the economy at that point in time. I also agree with the other commenter that it may be better to see some research prior to December 2019.

alex_young 7y ago
Are you implying that we can’t draw conclusions from limited data sets? Surely there must be some value in looking at the last event, even if it was 100 years ago...
- Nuzzerino 7y ago
  What value did you have in mind?
- tathougies 7y ago
  You have to think about confounding factors when considering only one event. The effects of the 1918 flu are completely overshadowed by the end of WWI. Thus you can't really draw many conclusions.
  - xapata 7y ago
    You can't control for that effect?
    - delecti 7y ago
      There's no control group for a single event.
      - itsdrewmiller 7y ago
        They are comparing different cities that both went through WWI ending, so I think it would require a reason to believe WWI would affect different cities differently to be a threat to validity. Certainly one could imagine there might be such reasons and it is worth discussing them, but presumably that would require looking at the specific cities in the paper which I presume no one here (including me) has done.
        tathougies 7y ago
        But it's not quite fair to compare two different cities in an economy where only some cities are shut down. If New York shuts down but can still purchase goods from LA, then that's a remarkably different situation than the entire country being shut down, which is essentially what is going on now. You cannot study the cities in isolation. They are not self-sufficient economies. It's possible that one city was able to isolate because it depended on the other cities to be working to receive its goods and to jump-start its economy after the isolation. These effects cannot be ignored, and are substantially different than today, when it's not just that most states are shut down but also that most countries are shut down (well the ones of economic significance, anyway).
        xapata 7y ago
        But if the effect of WWI across different cities is random, and not correlated with the explanatory variables, then the study can still work despite WWI affecting the target variable. It'll just be part of the error term.
      - xapata 7y ago
        No, the study used a cross-section, not a single observation.

alex_young 7y ago

Are you implying that we can’t draw conclusions from limited data sets? Surely there must be some value in looking at the last event, even if it was 100 years ago...

Nuzzerino 7y ago
What value did you have in mind?
tathougies 7y ago
You have to think about confounding factors when considering only one event. The effects of the 1918 flu are completely overshadowed by the end of WWI. Thus you can't really draw many conclusions.
- xapata 7y ago
  You can't control for that effect?
  - delecti 7y ago
    There's no control group for a single event.
    - itsdrewmiller 7y ago
      They are comparing different cities that both went through WWI ending, so I think it would require a reason to believe WWI would affect different cities differently to be a threat to validity. Certainly one could imagine there might be such reasons and it is worth discussing them, but presumably that would require looking at the specific cities in the paper which I presume no one here (including me) has done.
      - tathougies 7y ago
        But it's not quite fair to compare two different cities in an economy where only some cities are shut down. If New York shuts down but can still purchase goods from LA, then that's a remarkably different situation than the entire country being shut down, which is essentially what is going on now. You cannot study the cities in isolation. They are not self-sufficient economies. It's possible that one city was able to isolate because it depended on the other cities to be working to receive its goods and to jump-start its economy after the isolation. These effects cannot be ignored, and are substantially different than today, when it's not just that most states are shut down but also that most countries are shut down (well the ones of economic significance, anyway).
        xapata 7y ago
        But if the effect of WWI across different cities is random, and not correlated with the explanatory variables, then the study can still work despite WWI affecting the target variable. It'll just be part of the error term.
    - xapata 7y ago
      No, the study used a cross-section, not a single observation.