Last year, I wrote a blog post about how U.S. presidents have limited influence over the economy. As part of that post, I showed the relationship between the unemployment rate and participation in the Supplemental Nutrition Assistance Program (SNAP; formerly known as food stamps). That graph–a connected scatterplot–was, admittedly, a bit of a mess (though I thought the annotation was quite nice), but I didn’t want to do side-by-side line charts because the correlation is important.
So, what would you have done with this data? How would you have visualized it? I’m providing the data in Excel format as well as a downloadable (zipped) Tableau workbook.
I would consider using a small multiple approach, like the attached, so you can focus on particular years in each frame. I’ve not done this, but I would have your annotations per frame too.
Attached image:
I thought the most interesting way this graphic could be enhanced was through data manipulation. When considering a President’s impact on the economy, nothing he does will have an immediate impact. Legislation takes time to pass, and frequently does begin to have effects until years later. While a president being in office may correlate with particular economic events, these economic lags make assessing causation incredible difficult. In my visuals, presidential party is carried forward with the associated lag period. For example, a the 2010 point would be marked red for G.W. Bush on the 2-year lag chart.
Next steps:
Adding some nice looking annotations and interactivity.
The Python code for this analysis can be found here: http://localhost:8889/notebooks/visualizing_presidential_impact/Visualizing%20Presidential%20Influence.ipynb#
Attached image:
Sorry about that I posted the wrong link to the code and doesn’t look like I can edit. Public Link:
https://github.com/NickDoesData/visualizing_presidential_impact/blob/master/Visualizing%20Presidential%20Influence.ipynb
Isn’t the Obama label misplaced in the connected scatterplot? I would expect it to be on 2009.
Hi. The way i see it, i would plot unemployment/SNAP rates and population against time. A multi-level X-axis can be used to display the years, president and party in administration.
Attached image: