US Politicians’ Worth: A Wealth of Disclosures

If the citizen’s right to know shades into a gossipy prurience, the right to know nevertheless prevails; thus if the disclosures of the estimated net worth of American lawmakers compiled by the Center for Responsive Politics and reposed in the Enigma public data site titillates or outrages, one assumes the greater good is being served just the same. Those estimates are here:

LawmakerNetWorth-2014

First, don’t be fooled by the 2014 date stamp, as I was at first blush. After all, the worth list, sorted as it is highest-to-lowest, is headed by the irrepressibly well-heeled Mr. Trump, who if memory serves only pulled into the White House in 2017, whether he wanted to or not. The 2014, then, must point to the date of the estimates instead (and as such, the year field and its relentlessly uniform 2014s could be deleted, though keep in mind that Enigma stores Center worth data for previous years as well), and if a good many of the numbers appear to you in scientific-notational form (e.g. 1.07E+.08) that’s just one more way by which Excel means to tell you that their columns are presently too narrow. Just auto-fit away, as usual. (Note that the origin field in H is likewise uninformative, comprising nothing but the Center for Responsive Politics entry, and the source field in I is even sparser, amounting to nothing but its header.)

Get past those preliminaries and you’ll need to remind yourself about the codes informing the chamber field: E – Executive, S – Senate, H – House of Representatives, and J – Judiciary. That latter code estimates the worth of the nine now-serving Supreme Court Justices, an interesting set of disclosures about a coterie one might deem slightly above the financial fray; the S and H signifiers apparently name and enumerate each and every current Senator and Congressperson, a nicely comprehensive complement. The twelve executive entries, however, stand in a different relation to the dataset; they identify presidential hopefuls for the 2016 election and not office holders, with the exception of the winner.

Perhaps the first attention-grabber in the data to seize your lapels is the trio of worth estimates lined up for each politician. The Center for Responsive Politics explains that the data spring from disclosure forms completed by office holders which “do not require exact values”, but rather “the range of value into which an asset falls”. The Center thus filled ranges with lower and upper estimates and averaged the two into the avgvalue field, those computations in turn exhibiting considerable plus-minus variation. Indeed, 77 of the politicians declared a negative lower-end asset value, with another 11 submitting negative upper estimates. Of course, our data can’t take that question further.

In any case, working with the averages we could broach a first obvious question, namely how worth associates with chamber:

Rows: chamber

Values: avgvalue (formatted to two decimals with a comma)

I get:

Remember of course that the Executive universe is small, numbering but 12 cases. As such the 1.482- billion-dollar average estimate for Donald Trump slingshots the mean skyward, in spite of negative-asset claimers Scott Walker, Rick Santorum, and Martin O’Malley. Whether that means erstwhile Maryland governor O’Malley truly has less money than you and I is perhaps a query better put to his accountant, but the data maintain as much.

But the larger, if ingenuous, conclusion urged upon us is the reality that American national politicians – including Supreme Court justices – are overwhelmingly multi-millionaires. If the averages of the averages are credible, Senators possess over $10,000,000 worth of assets, with Representatives checking in a $7,500,000, accumulations that of course can’t ascribed to their $174,000 salaries. At the same time, however, variation within the Congress is massive, including outlier Darrel Issa, a Republican Congressman from southern California worth, per his average estimate, over $436,000,000.

For a sense of the dispersion of wealth, then, you can drag avgvalue in to values again, setting it to Stddevp (for the operational difference between Stddev and Stddevp, look here):

When the standard deviations exceed the averages you’re dealing with a lot of variation, even after discounting the ten-figured Mr. Trump.

Now for perhaps the juicier question: Might net worth vary palpably by party affiliation? Conventional wisdom would respond in the affirmative, i.e. Republicans and their putative sympathies for the haves could be expected to have more as well than the latter-day Blues. But either way, the dataset should enable an answer of sorts to the question.

The marching order here is to isolate party, as it’s referenced by its initial in the name field. Because that initial appears (almost) in every case immediately after the open parenthesis alongside each surname, we can take over the empty source field, rename it Party, and write in I2 (assuming you haven’t deleted the extraneous year field):

=MID(B2,FIND(“(“,B2)+1,1)

That expression seeks the character position of the open parenthesis and adds one to it, enabling MID to grab the next single character, which denotes political party. Copy the formula down the field and that should work, with nine exceptions – the Supreme Court justices, whose party memberships have been withheld from the data, an exclusion that imposes a #VALUE message upon their formulas. The workaround: rewrite the formulas thusly:

=IFERROR(MID(B2,FIND(“(“,B2)+1,1),”J”)

The nifty IFERROR function returns the default, nested formula result unless it evaluates to a error, in which case it’ll interject the defined if-false result, in the above case J, for Judiciary.

Once that revamped expression is copied down the field the pivot table follows:

Rows: Party

Values: avgvalue (Average, with appropriate formatting)

I get:

The thesis is confirmed, apparently (the I represents the two independent legislators informing the list, by the way, one of whom is Bernie Sanders). Republicans are worth on average 47% more than the (very) relatively proletarian Dems; but toss Stdevp back in to the table and we see:

The mammoth standard deviation among GOP sectarians has a good deal to do with the Trump billion, and Issa’s 436,000,000, too. Thus, if one deletes the row bearing Trump’s data and refreshes the pivot table, you’ll see:

Check it out.  The Republicans, we see, in actuality average a smaller net worth and standard deviation than Democrats. Go further and delete all 12 Executive entries – none of whom, with the exception of the incumbent, are election officials, after all – and run the pivot table once more:

Here, Democrats possess 9.94% more worth on average than Republicans.

Now that’s unconventionally wise.