Watch the 12-minute slideshow (audio included) below to get a better understanding of the state economic and fiscal outlook, including ideas for a balanced solution to the budget shortfall.
Click on the green “play” button on the bottom to begin the narrated slide show. The large black arrow on the right-hand side just skips forward to the second slide.
Note: if you cannot see the presentation try using the Firefox browser instead of Internet Explorer.
Click on the green “play” button on the bottom to begin the narrated slide show. The large black arrow on the right-hand side just skips forward to the second slide.
Note: if you cannot see the presentation try using the Firefox browser instead of Internet Explorer.
7 comments:
You seem to use a lot of non-zero starting points for your graphs. I presume you do this in order to add more drama to the depictions of job losses, unemployment increases, etc. But it's terribly misleading, and caters to the kind of statistical and economic illiteracy that does not serve your cause well.
I know your organization's work. I *know* that you understand how misleading it can be to show a graph with a bottom of 2.8 million jobs in order to make a loss of 130,000 jobs look like cliff diving. (And yes, the impact is large, but it's not as if 90% of jobs have been lost, which is what your graph seems to communicate.) So I urge you to consider whether it's helpful in the long run to concoct misleading graphs like this, or whether it's more helpful to find clear, precise, and accurate ways of communicating the data.
High-quality progressive information on the budget is a critical need. Doing your graphs USA Today-style does not serve that need. Please stop going down this road. I suggest attending an Edward Tufte seminar and really thinking about your long term vision of an engaged & economically literate community of decision markers, and whether or not your practices serve this goal. Thanks!
Using a non-zero baseline is not uncommon in statistical analysis and can be found in any scientific journal. In particular, it is accepted in a time-series analysis when the zero point does not naturally occur (zero jobs in the state?). The job loss that has occurred in Washington State is indeed a statistically significant and meaningful change. Insisting on a zero axis would hide the truth, not illuminate it.
Tufte does not take your hardline stance on this issue, by the way.
Thanks for putting this presentation together; I found it really informative.
And for the record your graphs are fine. :)
I want to also say thank you for putting this presentation together.
I have always been against raising taxes to make up for shortfalls as I have believed in the past that cutting the fat should outweigh implementing higher taxes to make up for it.
However, in the current state of the economy, I agree that removing services that would put our state in jeapordy of falling behind comparative to other states is not a sensible approach to our future.
Could you put the presentation in a printable form? Thanks.
Jeff: I 100% agree that non-zero baselines aren't always incorrect and aren't always misleading. But they are terrible misleading in this presentation. My issue is that the truth is way bad enough -- but your non-zero-baseline graphs suggest a sort of hysteria about the magnitude of the economic changes going on that simply don't map to anyone's short-term experience and therefore undermine your whole presentation. (Should we be more outraged about what's going on? I think we should, but we don't get there by starting from hysteria.)
I like you guys. Your research is great. Your data is great. But your presentation is just misleading. It's like if Tim Eyman drew a graph of the state budget that showed recent budget growth over the last decade as a skyrocketing, nearly vertical line. That would be easy enough for him to draw by choosing the right baseline and axis-scales. And yes, changing scales is a common statistical technique. But it would still be a terribly misleading piece of information propaganda, because these common and worthwhile tools would be used to coax towards wrong conclusions rather than to help understand. Please don't fall into that same trap. Progressives need to be better than that.
And the relevance of Tufte is not on non-zero baselines per se. It's about presenting rich information in a way that lets the information receiver re-create the process of drawing conclusions. Data as evidence, not propaganda.
Fat-tailed: feel free to contact us directly - through email or phone - if you'd like discuss your suggestions.
Post a Comment