Why 2010 Census Found Fewer Living in NYC and Long Island

Many analysts and government officials have been surprised by the lower than expected 2010 census population counts for Long Island and New York City, and indeed, the entire state of New York. In fact, the Census Bureau itself, in its annual Population Estimates program, and the Long Island Power Authority, in its annual Long Island Population Survey, appears to have overestimated the counties’ populations, most likely the result of a steeper than expected increase in the number of residential vacancies, and a steeper than expected decline in the number of illegal/unauthorized residents.  A new Long Island Regional Planning Council report (scroll down to “reports and documents”) makes the following points.

1.      In all likelihood, the Census Bureau’s Population Estimates program did not capture the sudden but widespread effect on the region’s housing and immigrant populations of the 2007-2010 economic down turn. That is, an increase in the number of housing vacancies, caused by home foreclosures and the departure of low-skilled, mostly undocumented immigrant laborers from abroad due to job declines, eluded the most commonly used survey and forecasting methods.

2.      This led to annual population estimates leading up to the 2010 census that raised anticipated population counts for the region.

3.      While the Census Bureau’s decennial census is not perfect and a full count of the nation’s population remains an elusive goal, it is a reliable measure of residential vacancies, and is the only survey that provides door to door canvassing.  

4.      For Long Island (Nassau and Suffolk counties) the Census Bureau’s 2009 population estimate of 2,875,904 was 43,022 (1.5%) higher than the actual 2010 census count of 2,832,882. In New York City, the Census Bureau population estimate for 2009 was 8,391,881, or 216,748 (2.7%) higher than the 2010 census count of 8,175,133.

5.      In the 2010 census, the Nassau-Suffolk region had a 2 percentage point increase in the proportion of total residential housing units that were vacant compared to 2000. This translates into a 26,054 increase in the number of vacant housing units for Long Island.

6.      Using conservative estimates, the 26,054 increase in vacant housing units would result in a loss in the number of residents of approximately 38,556. 

7.      The increases in annual foreclosure filings for 2007-20010, using 2006 as the base year, totaled 20,258 on Long Island.

8.      Seasonal homes, mostly located on the east end of Long Island, do not appear to account for the increase in housing vacancies. The ten years between 2000 and 2010 saw only 405 more seasonal homes built on Long Island’s east end than the ten years between 1990 and 2000.

9.      In New York City, the 2010 census found an increase of 81,954 in the number of vacant housing units since 2000. These additional vacant housing units are estimated to result in approximately 163,908 fewer residents of the city.

10.  The Census Bureau’s Population Estimates program, which doesn’t count housing units, appears to have failed to capture a decline in the  unauthorized immigrant population.

11.  The Census Bureau’s Population Estimates program, which like other Census programs, does not ask about the legal status of immigrants, assumed that net international migration remained relatively flat between 2007 and 2009 (2010 data is not available), showing a decline of just 931 for Long Island. For New York City, the estimated annual decline between 2007 and 2009 totaled 8,037.

12.  Using data supplied by the U.S. Department of Homeland Security, though, it is estimated that New York City’s unauthorized population shrank by 69,300 between 2007 and 2009, from 492,800 to 423,500.  For Long Island, the unauthorized population is estimated to have declined by 10,583 between 2007 and 2009, from 91,526 to 80,943.

13.  The conclusion to be drawn is that the economic down turn of 2007 to 2010 greatly reduced the ability of the most reliable population measures to capture rapid and steep population movements among “hard to count” populations.

14.  This conclusion is backed by a small but relevant example, the community of Montauk, in East Hampton, Long Island. The decline in total population in Montauk of 16%, or 525 people (3,326 vs. 3851) between 2000 and 2010, is almost precisely equal to the loss of Hispanic residents in Montauk of 537, a 41.3% decline between 2000 and 2010. The majority of low-skilled, possibly unauthorized migrant workers on Long Island’s east end are of Hispanic origin.