Should You Trust Any Numbers Coming Out Of The U.S. Government?

The big economic news today comes from the release yesterday by the Census Bureau of its annual report on Income and Poverty in the United States.  Data in the report are the basis for today's lead front page articles in both the New York Times ("Incomes In U.S. Are Up Sharply; Poor Gain Most") and the Wall Street Journal ("Family Incomes Rise After Lull").  

And the news -- if you believe it -- is indeed remarkably good.  Median household income is reported to have risen $2,798, or 5.2%, from $53,718 in 2014 to $56,516 in 2015..  The official "poverty rate," is reported to have declined from 14.8% in 2014 to 13.5% in 2015.  That's a full 1.3%, and represents some 3.5 million fewer people "in poverty" in 2015 compared to 2014.  And the biggest gains in income were at the lowest income levels: real household incomes are reported to have risen 7.9% for households at the 10th percentile of the income distribution, and 6.3% for those at the 20th percentile.

Of course, the report emerges in the heat of the presidential election campaign, where it cannot fail to be interpreted to support the candidacy of the candidate of party of the incumbent President.  From the Times:

The data was [sic] released into a heated presidential race, where Democrats seized on the statistics to promote Hillary Clinton's candidacy and undercut Donald J. Trump's dark assessment of the nation's well-being.

And President Obama himself took the opportunity to take credit at a speech in Philadelphia:

"We lifted three and a half million people out of poverty, the largest one-year drop in poverty since 1968," President Obama said on Tuesday at a rally in Philadelphia for Mrs. Clinton.  "The uninsured rate is the lowest since they began keeping records.  The pay gap between men and women shrank to the lowest level on record," he said, adding, "Thanks, Obama."

So is there any reason to be suspicious?

Well, start with this.  Any possible idea you might have that the bureaucrats at the Census Bureau are just neutral experts reporting the data fairly and accurately will be completely shattered if only you look into the matter of the "poverty rate" and how it is compiled and reported.  I have written about this too many times to count, and will not go into all the details here.  For more background and links, try this post, or this one.  But here's the summary:  The definitions and categories that underlie the "poverty rate" have been intentionally manipulated and gerrymandered to support the ongoing growth of failed government "anti-poverty" efforts and of government staff and budgets.  Over many years, definitions have been set to exclude most government benefits from the measure of poverty in order to keep the "poverty" figures high so that they can be used to advocate for yet more spending -- spending that then again will not be counted when "poverty" is measured.  And the blame for this cynical game can only be placed on the Census Bureau itself, which sets the definitions and parameters for collecting and reporting the data.  So, without a doubt, the Census Bureau is suspect.  It puts out numbers which are clearly and obviously manipulated to support advocacy for more and bigger government programs, and that support political candidates who will support and advocate for the same.

So now to today's numbers.  Median household income up 5.2%?  So, how did GDP do?  Real GDP was up 2.4% in 2015 over 2014, according to BEA here.  So, supposedly, median household income increased by more than double the increase in GDP?  That seems to stretch the limits of plausibility.

The Census report that came out yesterday compared 2015 data to 2014.  Let's take a look at the comparable Census report that came out at this time last year, comparing 2014 data to 2013.  Here it is.  As further background, the increase in GDP from 2013 to 2014 was reported to be exactly the same as the increase from 2014 to 2015 -- 2.4%, according to Forbes here.  And how did that 2.4% increase in GDP in 2014 get reflected in median household income and poverty?  According to last year's Census report:

  • Median household income was reported to be $53,657 in 2014, which was actually down from $54,462 in 2013.  (They do claim that the decrease was "not statistically significant.")
  • "In 2014, the official poverty rate was 14.8 percent. There were 46.7 million people in poverty. Neither the poverty rate nor the number of people in poverty were statistically different from the 2013 estimates."   

So, in summary, from 2013 to 2014, as reported in 2015, GDP went up 2.4%, and that translated into a small decrease in median household income, and no change at all the the poverty rate or number of people in poverty.  From 2014 to 2015, as reported yesterday, GDP went up the same 2.4%, and that translated into a 5.2% increase in median household income, and a nearly 10% decrease in the poverty rate and number of people in "poverty."  Do you believe it?  The critical difference between the two years is, of course, that in 2015 the administration needed low numbers for income and high numbers for poverty in order to advocate before Congress and the people for more government spending and programs, while in 2016 the administration needed higher numbers for income and lower numbers for poverty in order to support the candidacy of Hillary for President.

While I'm at this, I should remind you of a few other relevant things.  The Census Bureau has long been a part of the Department of Commerce.  In 2009, after being elected President, and in preparation for the 2010 decennial census, Obama caused the Census Bureau to change its reporting to report to the White House directly.  Many at the time (e.g., John Fund here) were critical of that move as undermining the "independence" of the Census Bureau (although I was dubious that they ever had any real independence).  Fast forward to today, and I can't say what the practical reporting lines of the Census Bureau may be, whether to the Secretary of Commerce or to the White House directly.  However, the Secretary of Commerce is one Penny Pritzker.  She made her fortune selling your confidential personal information without your permission as head of the Trans Union credit bureau, and followed that by serving as finance chair of Obama's campaigns.  So really, is there any conceivable reason why you would trust anything coming out of the Census Bureau?  By the way, both the Times and the Journal report yesterday's numbers without the slightest trace of skepticism.  Hey, it's the government!  Of course they are the fair and neutral experts!

UPDATE, September 15:  John Crudele of the New York Post -- a long time critic of the Census Bureau -- has gone through the methodology, and thinks he has discovered how they came up with this year's numbers:

Census moved the goal posts. Starting in 2013 with a partial phase-in, which was fully implemented in 2014, Census changed the questions and the methods in calculating household income.  For example, Census, starting in 2014, began to “collect the value of assets that generate income if the respondent is unsure of the income generated.”  Also, the government started to use “income ranges” as a follow-up for “don’t know” or “refused” answers on income-amount questions.  There are plenty of other changes — but with just these two, income levels reported could be noticeably higher, say 5.2 percent higher, without the actual income being 5.2 percent higher.  In the fine print, Census admits the change. “The data for 2013 and beyond reflect the implementation of the redesigned income questions.”  Americans, in their guts, know the 5.2 percent gain in median household incomes isn’t true.  Now they know why they should follow their gut.

Now, when Census cooked up these methodology changes just in time for their pre-election income and poverty report, did they know that the changes would drive a nice apparent income increase and poverty decrease?  Of course they did.  Unfortunately, there is nothing honest about this.