Correlation does not always equal causation
This graph shows that the more people drive, the more they park and generate revenue for parking lot and garage operators. While there’s clearly a correlation between these two indicators, it isn’t clear that there’s a straightforward causality between them. In fact, a third indicator may be affecting the other two: the number of cars in use, the size of the road network, economic activity in general, commuting distance… Or maybe it’s a combination of all or some of these. This ambiguity is what makes statistical analysis much more complex than simply looking at correlations in a graph. FRED helps you stay rigorous by allowing you to download data into your favorite statistical software, either with a download from FRED itself (for example, via the “Download Data” link below the graph) or natively from the software of your choice. For starters, you can use this published data list.
How this graph was created: Search for and select “parking lot revenue” and click on “Add to Graph.” From the “Edit Graph” menu, search for “GDP deflator” in the “Customize data” section and add the series, applying formula a/b. Then from the “Add Line” tab, search for and add “vehicle miles.” Finally, from the “Format” tab, place the y-axis of the second line on the right side.
How this data list was created. For starters, you need to (create and) log on to a FRED account. Then, from any account page, click on “Add new” and select “Data list.” Give it a name. Then search for the series, check the series you want, and click on “Add to data list.” Repeat until satisfied. You can make the data list public and will be required to give it a public name.
Suggested by Christian Zimmermann.
View on FRED, series used in this post: