3 ms·

Very misleading Y axis. Here is the same plot with a proper axis: http://i.imgur.com/k4ClDSF.png http://i.imgur.com/k4ClDSF.png

pessimizer 12y ago
That one is the misleading one. We're interested in the magnitude of the change vs. the recovery, not the numbers 0 to 100. A graph of GDP plotted as a percentage of current GDP at that scale could make the recession seem like a tiny little 2% blip.
- nostrademons 12y ago
  The recession is a tiny little 2% blip, if you're one of the ~85% of people who were not laid off and didn't have any trouble finding a job. And it's a very large 2% blip if you're one of the people who are. My point is that perspective matters - a lot. By the numbers, this recession is worse than any since the Great Depression. By the numbers, this recession is only a small percentage of the total U.S. economy. Which numbers are correct? Well, they're actually the same numbers, what matters is how you use them and what conclusion you're trying to draw. From my personal perspective, the recession was great. It meant I could actually get decent rents in the Bay Area for the first two years after I moved out here, and I could pick up stocks for relatively cheap, and I had no problem getting a job. I realize this is not the perspective of many other people, and the numbers help tell me how many other people. And one of my professional startup interests is finding better ways to help people manage their careers, so I'm quite interested in hearing other people's perspectives (if you have stories or want to vent, feel free to e-mail me...I'm quite happy to listen, my e-mail address is in my profile.) But understand that numbers always require interpretation - whether the graph is misleading or not depends entirely on where you're leading people.
  - pessimizer 12y ago
    Agreed. My point is that the data that we're examining are the changes in employment to population surrounding the housing bubble. Measuring on a 0% to 100% scale whether we've begun a recovery from a recent drop in the rate of employment is like measuring the tides from the sea floor. Personally, my salary has tripled since the recession, and my wealth increased by a factor of five or six. The adult, established upper-middle class is doing fine. My pre-tax savings rate is over 50% though, so I'm not doing a consumption-driven economy any favors.
  - sillysaurus3 12y ago
    Very good points! From my perspective, the recession was devastating. Not for me, but for a friend's family. One of them owned a side business involving construction. Not enough to pay all the bills, but in '07 suddenly it was. So he quit his regular full-time job, even though he had a full family to support. It wasn't a stupid decision though, because he had so many contracts (guaranteed cash flows) that he could support his whole family twice over. I remember that business was so good that he made a big show of unveiling a wonderful grand piano for the whole family. He had the contracts to support such a lifestyle. Then '08 happened, and the contracts were all canceled. (Illegally, but that's beside the point; he couldn't afford to pay any bills, let alone a lawyer to go after anybody.) I later found out he had contemplated an elaborate suicide to collect on death insurance to provide for his own family. Not trying to contradict anything you've said; merely adding an alternate perspective.
- DangerousPie 12y ago
  I was mostly thinking about the "it illustrates how impactful the '08 crash was" in the original comment. Of course it looks like a huge change if you restrict the Y-axis, but if you plot it on the full axis you realize that it was actually still relatively small. It did of course affect millions of Americans and might have been the biggest decrease in a long time, but it was nowhere near the 80% drop that it looked like in the original plot.
malandrew 12y ago
Neither are really misleading or particularly enlightening by themself. In this particular case the Y-axis alone is not enough to draw conclusions because we don't know what magnitudes of movement are normal, what are healthy and what are unhealthy. Choosing 0 to 100 doesn't add anything because what does it even mean to have 0% employment or 100% employment. Those are not even reasonable bounds to start with given the topic at hand. I can't even imagine any conditions under which either extreme is plausible. A peers comp here would be far more illustrative than choosing any particular Y-axis scale. If for example the same employment figures for all other countries were plotted on the same graph, we would see extremes, from the country with the lowest employment rates to the highest employment rates. Those figures would serve to provide reasonable bookends for a graph. If no economy in the world has an employment rate of less than 45%, then a lower bound of 40% would be reasonable. If no country exceeds 85%, than an upper bound of 90% would be reasonable. The scale chosen should illustrate plausible ranges of employment for any modern economy. What does matter in all this is the time series plotting and whether or not this data suggests that there may be a virtuous/vicious feed back loop when the employment deviates from some stable level. For example, once employment falls below a certain point, does that contribute to a perception that businesses should be tightening their belts, further exacerbating the problem. Likewise, when the figures are moving up, businesses anticipate economic growth and therefor start hiring in preparation for the growth. This graph is the equivalent of measuring distance, when measuring the first (velocity) or second (acceleration) derivative might be more illustrative of how serious all this is. When matters is not how much something has fallen or rising, but how fast it is falling or rising and if it is starting to speed up or slow down.

pessimizer 12y ago

That one is the misleading one. We're interested in the magnitude of the change vs. the recovery, not the numbers 0 to 100. A graph of GDP plotted as a percentage of current GDP at that scale could make the recession seem like a tiny little 2% blip.

nostrademons 12y ago
The recession is a tiny little 2% blip, if you're one of the ~85% of people who were not laid off and didn't have any trouble finding a job. And it's a very large 2% blip if you're one of the people who are. My point is that perspective matters - a lot. By the numbers, this recession is worse than any since the Great Depression. By the numbers, this recession is only a small percentage of the total U.S. economy. Which numbers are correct? Well, they're actually the same numbers, what matters is how you use them and what conclusion you're trying to draw. From my personal perspective, the recession was great. It meant I could actually get decent rents in the Bay Area for the first two years after I moved out here, and I could pick up stocks for relatively cheap, and I had no problem getting a job. I realize this is not the perspective of many other people, and the numbers help tell me how many other people. And one of my professional startup interests is finding better ways to help people manage their careers, so I'm quite interested in hearing other people's perspectives (if you have stories or want to vent, feel free to e-mail me...I'm quite happy to listen, my e-mail address is in my profile.) But understand that numbers always require interpretation - whether the graph is misleading or not depends entirely on where you're leading people.
- pessimizer 12y ago
  Agreed. My point is that the data that we're examining are the changes in employment to population surrounding the housing bubble. Measuring on a 0% to 100% scale whether we've begun a recovery from a recent drop in the rate of employment is like measuring the tides from the sea floor. Personally, my salary has tripled since the recession, and my wealth increased by a factor of five or six. The adult, established upper-middle class is doing fine. My pre-tax savings rate is over 50% though, so I'm not doing a consumption-driven economy any favors.
- sillysaurus3 12y ago
  Very good points! From my perspective, the recession was devastating. Not for me, but for a friend's family. One of them owned a side business involving construction. Not enough to pay all the bills, but in '07 suddenly it was. So he quit his regular full-time job, even though he had a full family to support. It wasn't a stupid decision though, because he had so many contracts (guaranteed cash flows) that he could support his whole family twice over. I remember that business was so good that he made a big show of unveiling a wonderful grand piano for the whole family. He had the contracts to support such a lifestyle. Then '08 happened, and the contracts were all canceled. (Illegally, but that's beside the point; he couldn't afford to pay any bills, let alone a lawyer to go after anybody.) I later found out he had contemplated an elaborate suicide to collect on death insurance to provide for his own family. Not trying to contradict anything you've said; merely adding an alternate perspective.
DangerousPie 12y ago
I was mostly thinking about the "it illustrates how impactful the '08 crash was" in the original comment. Of course it looks like a huge change if you restrict the Y-axis, but if you plot it on the full axis you realize that it was actually still relatively small. It did of course affect millions of Americans and might have been the biggest decrease in a long time, but it was nowhere near the 80% drop that it looked like in the original plot.

malandrew 12y ago

Neither are really misleading or particularly enlightening by themself. In this particular case the Y-axis alone is not enough to draw conclusions because we don't know what magnitudes of movement are normal, what are healthy and what are unhealthy. Choosing 0 to 100 doesn't add anything because what does it even mean to have 0% employment or 100% employment. Those are not even reasonable bounds to start with given the topic at hand. I can't even imagine any conditions under which either extreme is plausible. A peers comp here would be far more illustrative than choosing any particular Y-axis scale. If for example the same employment figures for all other countries were plotted on the same graph, we would see extremes, from the country with the lowest employment rates to the highest employment rates. Those figures would serve to provide reasonable bookends for a graph. If no economy in the world has an employment rate of less than 45%, then a lower bound of 40% would be reasonable. If no country exceeds 85%, than an upper bound of 90% would be reasonable. The scale chosen should illustrate plausible ranges of employment for any modern economy. What does matter in all this is the time series plotting and whether or not this data suggests that there may be a virtuous/vicious feed back loop when the employment deviates from some stable level. For example, once employment falls below a certain point, does that contribute to a perception that businesses should be tightening their belts, further exacerbating the problem. Likewise, when the figures are moving up, businesses anticipate economic growth and therefor start hiring in preparation for the growth. This graph is the equivalent of measuring distance, when measuring the first (velocity) or second (acceleration) derivative might be more illustrative of how serious all this is. When matters is not how much something has fallen or rising, but how fast it is falling or rising and if it is starting to speed up or slow down.