US Census

How Census Data Gets Weaponized (Even When the Data Isn’t Actually Great)

Bending the stats to fit a narrative (abcnews)

Bending the stats to fit a narrative (abcnews)

This week, the US Census reported some generally good news economically.  For the first time since 1999, wages across all income levels have increased, the poverty level has decreased, and the number of uninsured has also decreased. This is undoubtedly welcome news for many households.  However, a few people, including Kriston Capps over at City Lab, have pointed out that these numbers don’t tell the full story.  The cost of housing looms large over this discussion, deflating the positives in this story considerably.  That doesn’t mean the report won’t become a useful tool in competing political narratives - while ignoring the broader picture.

Let’s start with the trends.  The US Census Report estimated that between 2014 and 2015, the real median household income increased by 5.2%.  During that time period, the official poverty rate fell by 1.2% to 14.3%.  Finally, the number of uninsured fell by 1.3%, resulting in 90.9% of Americans having some form of coverage. These are clearly welcome developments.

There are many ways to measure income, poverty, and insurance coverage. Inevitably certain assumptions are made that include or exclude variables that would change these numbers.  For example, the US Census report described in detail the difference between the official national poverty level and the supplemental poverty level. Without going into the details, it basically comes down to how they factor in children under-15 in a household.  I’m personally not sure how you could justify not counting them, but I won’t get into the nuances of this argument today.

The point is to say that these numbers don’t give an entirely accurate (or at least detailed) picture of the state of the economy and no one at the Census or any policy wonk would make that claim.  

A more detailed (and problematic) view of the numbers becomes apparent when you look at the geographical picture for more context.  The simple truth is that most of the income gains came in major cities (+7.3%), some suburbs around these cities (+4%), but none in rural areas. They actually saw a decline in income (-2%.)**

[**Author's Note: The day after publishing this blog, an article appeared that showed that the Census information on rural income was incorrect. It has actually increased by 3.4%. The discrepancy is a result of the definition of 'rural' which is slightly different between the Current Population Survey (cited in this blog) and the more finely-detailed American Community Survey, where the more accurate number comes from.]

I’ve written about why most economic gains are happening in increasingly fewer parts of the country, mostly concentrated around coastal cities and a few in-land cities.  Those cities are where the good jobs are.  It’s also where the ‘bad’ jobs are.

This fact alone might not be a bad thing, although you would hope for more diversity in job creation regionally from perhaps a cultural and political standpoint. The problem is how hard it is for people to move to these cities where those good or bad jobs are - and how hard it is to afford to remain living there.  Mobility is not following productivity or vice versa as pervious economic expansions have tended to do.  It’s making it harder for people (especially the poor) to actually gain from the overall economic expansion. That's not a sustainable model.

It’s no mystery why this is true – rent in these cities is expensive and rising.  That’s why the wage increase numbers aren’t really that impressive.  Sure, wages went up in cities 7.3% in 2014-2015 but rental costs went up nationally during that period by 4.6%. The gains are more than wiped out without even looking at the huge variances between cities.  

So even though the basic facts of this report are all positive, the details show a much more fragile and structurally broken landscape economically.  We see this disconnect when we look at recent polls. 75% of Americans are concerned about losing their homes in a downturn and over 50% of Americans are cost burdened (paying more than 30% of their income on housing). 

If all of these people are that anxious and insecure about their housing status, how is the cost of housing not a bigger issue politically?

Housing isn’t a major political issue because it doesn’t fit the narrow set of narratives identified by the media.  This is generally true, but especially true in a presidential election year. There’s the usual horse race coverage that the media loves but doesn’t ever actually address policy discussions. There’s the ‘mood of the nation’ coverage that vaguely taps into the public interest but rarely gets into deeper issues and concerns. And then there is this year’s hyper-partisanship coupled with the bizarro nature of Trump v Clinton that has all but drowned out meaningful policy exchanges.

So instead of supplementing this Census report with important context about what these numbers say and what they don’t say (and how they fit in to the larger political and economic landscape), media coverage has folded it into three predictable narratives: the apolitical and shallow: “Everyone is Doing Great Now in Economy;” the partisan spin: “Hilary Deserves Obama Third Term;” or the partisan rejection: “Economy is Doing Terribly Under Obama/Clinton.” 

Whether it’s a jubilant Jason Furman in the Obama Administration or a self-assured Paul Krugman at the NY Times, you can see the Pro-Hilary spin.  Whether it’s a “surprised” Brietbart or a perma-skeptical Fox News, you see the Trump/Republican spin. One side plays up the shallow headline reading of the report, the other downplays or attempts to question it.  Neither side spends anytime in the nuance of the actual report and no one from the political media class appears to be either.  Most of the media just covers the headline.  That’s not a formula for policy action.

None of these narratives are helpful or even ‘true’ in a useful sense.  We should of course be glad that income is rising and poverty and uninsured rates are falling, but is the economy actually ‘healthy?’ If the gains are all happening in places where the cost of living is increasing significantly, and the losses are all concentrated in rural areas, shouldn’t this be alarming? Shouldn’t this prompt actual discussions about how to alleviate the housing crisis and create more economic opportunity?

It’s not the job of the US Census to provide broader context to their reports. It’s certainly not their job to provide policy descriptions (they aren’t even allowed to set their own definitions for what or how to measure something).  But if we can’t rely on our elected officials or the media to contextualize this information and to ask the right questions about what it means, how can we even begin to address the larger challenges facing our country? 

The New York Times Doesn't Get the Housing Market

Wait, seriously? (newyorktimes)

Wait, seriously? (newyorktimes)

This week the New York Times published an article about how the national housing market “finally looks healthy” that was at best misleading and at worst irresponsible (it was also lazily edited).  The writer bases his argument on a joint release from the US Census and HUD that shows an increase in housing sales of 12% over last month and 31% over the same month last year.  On the surface, this could be a positive sign that the housing market is returning to ‘health’, but once you dig into the numbers, the picture is more complicated.  In fact, if this writer stepped back and examined the broader context of the economy, it would be hard to argue that we have a healthy market at all. This reveals a larger problem about how the media talks about the housing market and what it hasn’t learned from the 2008 crash.

Let’s start with the data from the joint report.  Though the article does point out that this data has a wide margin of error and is volatile, it still proceeds to use it as the core of the argument.  The problem is, this data does have a wide margin of error and is incredibly volatile because it’s based on a quick sample turnaround. You only need to look at the previous month’s release (June 2016) to see the difference in the July report’s revised totals.  They lowered the June numbers by 10,000 homes.   The report itself is clear about expectations and explains its methods and possible errors.  It also makes it clear that it takes a quarter to get an accurate sense of the trends within the market.  You can quibble with how different the data ends up being from what is currently reported, but the point is, this is a thin piece of evidence to base an entire argument on. 

Now, let’s just assume that the data is more or less accurate. We are still left with a more complicated picture than the article suggests, particularly when trying to compare the differences among regions.  Though the numbers are broken down into 4 regions, the raw numbers don’t show how wildly different the housing market is, even in certain parts of the same region.  For example, prices and availability for housing in and near NYC remain a competitive death-match while parts of Connecticut less than an hour from New York are struggling.   Does that mean the housing market is healthy?

The reality is that economic growth is happening in an increasingly smaller number of cities (and it’s almost exclusively cities) in certain pockets of the country that also have extremely expensive housing markets.  This is because of a limit in the supply of housing in these markets relative to the demand, whether as a result of geographical or political limitations. In either case, this situation is making it harder for middle class workers, let alone poorer workers, to be able to locate in these hot job markets. Hot job markets should create a housing boom, but that’s not what’s happening and it’s hurting our productivity as a nation.  That’s not the sign of a healthy housing market.  

This article also somehow neglects to mention (which is surprising given that the Times has covered this issue in another section recently) who is buying up a lot of these houses in certain market segments.  The entry of private equity firms into the housing market has been a quiet, powerful force building over the last 10 years.  Though the number of houses under control by private equity on a national level is relatively minor, the concentration of their portfolios in certain markets has had a huge impact on the perception of the housing market as a whole.  In many cases, they are buying up homes for straight cash.  They then rent them until the market reaches a certain point to flip them.  This speculation is perfectly legal, but it distorts the actual activity of house sales.

Also unaccountably left unsaid is the question over the financial health of the housing market.  No, another crash is not just around the corner, but the continuing receivership of Freddie Mac and Fannie Mae (who collectively back over 60% of US mortgages) has left the previous crash in an unresolved, frozen state.  Though many of the excesses that led to the crisis have been removed, the flawed assumptions underlying the housing market – that homeownership is a good political policy, that the government should support it as an economic policy - remain in place.  The fact that this policy costs taxpayers $150 billion a year while still failing to provide adequate housing for millions of Americans should challenge the assertion that this is a healthy market.  It’s not even a true market.

It is a mystery how the same paper that has addressed all of these issues at various times could allow an article to be published that breezes passed them all.  Maybe it’s just laziness or someone having an off-day, but it could also be someone pushing a narrative in search of story.   The media primarily talks about the housing market as an economic trend.  In doing so, it internalizes the assumptions of people who profit from the housing market regardless of how that process impacts the larger society and economy.  The housing market looked healthy, even robust, in the lead up to the 2008 crash, but underneath the sales numbers were real problems with earnings, financial instruments, and exploitation that went underreported until it was too late.  It is stunning and ultimately disappointing to see the Times revert back to this laziness and water-carrying.

It’s easy to cherry pick data from a report or to draw phantom conclusions from it, but it’s more important to report the full context of an issue to readers.  The housing market is not healthy.  Arguably it has never been healthy.  Trying to present it that way does a terrible disservice to the people that are suffering from it and to the people that are trying to solve it.