Joel Kotkin is an insightful writer on topics relating to demographics, and he has written much that is very interesting and informative.  A few days ago he had a piece in the Daily News titled "RIP, NYC's middle class: Why families are being pushed away from the city."  There is a lot to like about this article, but I also think that on many issues Kotkin is being seriously misled by useless and often fraudulent government data.

The thesis here is that while the rich in New York are getting richer, the number of poor is actually increasing and the middle class is stagnating.

As the cost of living has skyrocketed while pay has stagnated except for those at the very top, New York has shifted from a place people go to make it to a place for those who already have it made, or whose families have.  And once here, the rich are indeed getting richer even as the rest of the city is barely holding on.

The problem here is that Kotkin is relying on government data.  It's hard to blame him for that, because it's all we have.  But you really can't try to draw many conclusions from this stuff, at least not without seriously accounting for the obvious inadequacies and outright fraud practiced by the government in compiling the numbers.  When it comes to income, poverty, and income inequality numbers, there is good reason to think that the government has pretty good information as to the top two quintiles (probably less so as to the top 1%), but in the bottom three quintiles there are huge problems, and it gets worse the lower you go.  By the time you get to the bottom quintile, the government numbers are almost completely fake, failing to take account of around a trillion dollars a year in government handouts and completely missing unknown hundreds of billions, or possibly trillions, of dollars of off-the-books and illegal income.

So here is Kotkin's key piece of data:

Between 1990 and 2010, the city’s 1% saw their median income shoot up from $452,415 to $716,625 in 2010 dollars, even as the bottom 60% hardly saw their incomes budge at all, according to a recent City University study.

Kotkin doesn't cite the specific City University study, but it's not too hard to find information from Census data strongly supporting the narrative of the supposed "stagnation" of New York's middle class.  For example, here is city-wide and neighborhood-by-neighborhood data compiled by WNYC from Census information on New York median household income.  According to this compilation, the median went down from $54,057 in 2007-09 to $50,711 in 2010-12 (adjusted for inflation).  That sounds rather bad!  (Although note that among a small number of neighborhoods where median income increased, by more than 5%, two are Central Harlem and Bedford-Stuyvesant, New York's two most iconic African American neighborhoods.) 

But the question is, how much of the actual income of people in the bottom three quintiles is being captured by Census, and has the uncounted portion increased over time?  The two big categories of uncounted income are government in-kind handouts and off-the-books and illegal income.  The in-kind handout category has clearly soared over twenty years, and particularly during the Obama years.  Food stamp recipients nationwide have increased from about 30 million to almost 50 million just during the Obama years, meaning that there are now large numbers of recipients even in the second income quintile.  Medicaid has also greatly increased.  And now Obamacare is a whole new category of uncounted in-kind handout coming with its own big new incentives to minimize and/or hide at least some income in order to qualify for subsidies. 

How about that middle quintile, those just above and below the approximately $50,000 median?  There used to be lots of manufacturing jobs in New York that were the bread and butter for the people right around the middle.  Now there are hardly any manufacturing jobs.  Those manufacturing workers generally went to work at the same place for substantial companies every day, and their income was easily monitored by the government.  Today it's different.  For example, where I live, twenty years ago we had a huge rent-regulated housing stock in which no one had invested a dime in fifty years.  The buildings were declining and decaying.  The last 15 or so years have seen a very gradual phase out of rent regulation, together with a big run-up in housing values.  The result is that literally every second building has some kind of major construction or renovation job going on.  There are armies of people who work a few days or weeks on such jobs, often for small sub-contracting firms, before moving on to the next job.  Now, how much of that income does the Census Bureau capture?  In considering the answer to that question, take into account the tremendous incentives in today's handout society to show visible income below certain thresholds, whether it's to qualify for food stamps, or Medicaid, or Obamacare subsidies, or lots of other things.  

And that's just one example.  On the female side of the employment picture, the numbers of housekeepers and nannies are also vast.  Based on the market in my neighborhood, those jobs can pay around the median of the income distribution (and the median is for households that often include more than one worker).  Again, how much of that income is captured in the official numbers?  I only have anecdotal evidence, but frankly I think my anecdotal evidence is better in this instance than the Census data because the Census people completely know that they are missing large categories and they systematically report data missing those categories because they find it advantageous to support their agenda of advocating for more government spending.  Based on reports from my friends and acquaintances, well less than half of the income of nannies and housekeepers gets into official government information.  In fact, the government makes it literally punitive and almost impossible to be fully compliant in putting household workers on the books.  You must withhold federal, and state, and city, and FICA taxes; and you must pay workers compensation insurance, and unemployment insurance, and disability insurance.  And each of these must be paid to different entities and with different schedules and deadlines.  And once you're on their radar screen they audit you and torture you.  Or you can just pay cash.

Now if you ask the Census Bureau (which I did a few weeks ago) they will say that the recipients of their surveys are supposed to report all of this income, and don't worry, because we'll never tell the IRS about it.  Well, in the age of the Surveillance State, would you trust them to honor that promise?  You would be out of your mind.

People are not dumb.  Today, many are flocking to New York City to take jobs in the middle of the income distribution.  They would not be doing it if they saw no opportunity.  The fact is, it's the government numbers that are wrong.  I strongly suspect that the state of New York's middle class is just fine.  But every year, in the attempt to achieve perfect fairness, we put into place yet a few more incentives to minimize reported income.   So the numbers show stagnation or decline.  There is no reason to believe them.