NEW: Section 9:  Latest One-Year and Multi-Year Updates through 2008
                                   Section 8:   The "well off" population in The Bronx

DISCOVERING THE BRONX:

USING CENSUS DATA TO HIGHLIGHT SOCIAL PROBLEMS AND ACHIEVEMENTS IN A MAJOR URBAN AREA


TABLE OF CONTENTS:   VIEWING THE BRONX FROM DIFFERENT PERSPECTIVES

                    Most of the sections below include dozens of maps, graphs, charts, and explanatory text.
                    For these sections, only a couple of illustrative maps will be found on this page.
                    You must click on the hyperlink to see the whole section.

Introduction.  The 2000 Census and Beyond.  Special qualities of The Bronx:  part of New York City, but also a county in its own right.

Section 1.  First of all, Census data allows us to compare Bronx County with many other counties in the US (of course, many of the 3141 US counties have populations too small for meaningful comparisons).  In this national perspective,  certain ethnic, economic, and social indicators for The Bronx are striking, even unique.

Section 2.  Next we look at The Bronx in the New York City area - that is, compared to the rest of New York City and nearby suburbs.  Here the most useful geography is the zipcode - large enough to return reliable data, but small enough to show the diversity within this region.  Subjects considered include:
                                Racial and ethnic differences
                                    The Labor Force
                                    Education and other social indicators
                                    Poverty
                                    Over 200  items from Census 2000 for NY State, NY City, and The Bronx
                                       Excel file with these same 200 items for each NY metro zipcode

Section 3.  We then examine The Bronx as a whole, as it changes through time, using census data from as far back as 1960.
                                Detailed race and ethnic changes
                                    Changes in comparative size of age groups
                                    Changes in numbers of foreign born and in language usage
                                    Changes in family structure
                                    Changes in education levels and employment patterns
                                    Poverty level changes
                                    Housing and mobility changes
                                Detailed statistics covering Bronx changes, 1960-2000
 

Section 4.  Then we look at neighborhoods and smaller areas inside The Bronx, using data from census tracts (over 350), block groups (over 950) or even, when available, from  each city block in The Bronx (over 3500).  The focus is on racial-ethnic  stability and change within Bronx neighborhoods.   We also consider the ambiguities in precisely defining racial and ethnic groups.
                                Ambiguities and changes in the notion of race
                                Alternative ways to consider multiracial identification
                                The "official US government" racial categorization of Hispanics
                                Ethnic ebbs and flows, 1990-2000
                                    Distribution of major ethnic groups
                                    "New races" and multiracial identities
                                    unusual ethnic patterns for Bronx kids

Section 5.  For certain items, particularly race and ethnicity, it is useful to focus on the 2200 census tracts in New York City, to see patterns in each NYC neighborhood.  In this perspective we can better assess the uniqueness of many Bronx neighborhoods.

Section 6.  We then consider with charts and analysis how "normal" The Bronx is by comparing it with neighboring counties and with  larger geographic units (New York City; New York State; the US as a whole).  Is it useful  or confusing or even legitimate to do this?

Section 7.  We focus in more detail on changing social and economic realities in The Bronx.  There are three subsections:
                                 Sec.7a. Comparisions with other boroughs;
                                  Sec.7b. Detailed analysis of the relation of race to poverty;
                                 Sec.7c.  Recent patterns of migration into and out of The Bronx.

Section 8.  The complexity of The Bronx can be seen in the surprisingly large population of  well off, even affluent Bronx residents.  Comparisons to the whole Bronx population, and to NY City, NY State, and the US as a whole.
                                    Age and gender
                                    Race and ethnicity
                                    Family structure
                                    Education
                                    Language, citizenship, immigration
                                    Housing and transportation
                                    Population mobility

Section 9.  Most recent population updates through 2008:  Completeness and reliability of  Bronx data in an era of budget stringency.  The Bronx compared to the other NYC Boroughs and to other counties in the US.  Comparisons of PUMAs: opportunities and problems.

Afterthought.  How reliable are census figures?

At the end of this website we refer readers to our study of The Bronx in 1990.  Those interested in antiques may want to have a look.
 

INTRODUCTION:  THE BRONX, THE 2000 CENSUS, AND BEYOND


 

Welcome to The Bronx, New York City's only foothold on the North American mainland (the other parts of the city - Manhattan, Brooklyn, Queens, and Staten Island - are all on islands).  The Bronx, one of the five "boroughs" of New York City, is also a county in its own right.  Indeed, with close to 1.4 million people, The Bronx is the 26th largest county in the United States.

In area, The Bronx is among the smallest US counties.  Only Manhattan, Bristol, RI, and Arlington, VA (plus 31 of Virginia's independent cities) are smaller.  But, like Manhattan, The Bronx is just one part of a very large city - a part with a disproportionately large poor and working class population, while most of New York City's power and wealth lie in the southern half of Manhattan.  So, comparisons with larger and more diverse counties must be interpreted with care.

THE BRONX,  CENSUS 2000, AND MORE RECENT CENSUS DATA

At Lehman College, the only public four-year college in The Bronx, we in the Bronx Data Center have developed this website so we can use the extraordinarily rich details of the U.S. census to document the often unique diversity, problems and opportunities in The Bronx.   Through the internet, the Census Bureau has provided unparalleled access to all parts of Census 2000, and has been efficiently disseminating  updates on the national, state,  county, and now PUMA  levels.  We look forward to more precise intercensus data now that data for three successive years are aggregated for small areas like PUMAs.  But for the time being, Census 2000 data remains the most reliable basis for studying census tracts, blocks, and other small areas in The Bronx and elsewhere.

Obviously, studying only census data cannot lead to a complete understanding of Bronx society, history, art, sports, recreation, and so forth.  References at the end of the website point to analyses of The Bronx from other perspectives.

<>We now have many sources of Bronx census data for 2000 and more recent years.  The most useful materials for us are the following:
   ** The Census 2000 "short form" 100% census count, down to the level of block  or (for some files) census tract.  This file is generally referred to as "SF1." Released June, 2001
   ** The Census 2000 "long form" ("SF3").   Socio-economic data down to the block group level.   Released August, 2002.
   **  Yearly Population Estimates down to the county level provide the latest information on total population and its components (births, deaths, internal migration, international migration).   
These estimates are statistical manipulations based on 2000 data, so they are another way to substantiate or question the ACS estimates.  And they may indicate important trends.  Detailed statistics for all counties and States can be found at the Pop.Est. website, which is accessed through the Census Bureau's Factfinder website: http://factfinder.census.gov      The latest estimates show that the Bronx population in mid-2008 was 1,391,903, a gain of 4.4% since 2000.
   ** Starting in the late 1990's a new system was designed, ultimately to replace SF3.  It is called the American Community Survey (ACS), and has been gradually expanded  until in 2008 it finally covered most of the major geographic areas of the US (areas of at least 20,000 population).   ACS is based on monthly surveys of a sample of households - a much different methodology than the rest of the census, and therefore an interesting way to verify and even challenge official 2000 census data and updates such as recent Population Estimates.  Fortunately for us, The Bronx was one of 36 counties in the US participating in the trial runs of ACS from 1999 to 2007.  ACS has in the past recognized flaws in its methodology, especially with reference to The Bronx.  Indeed, there are also criticisms of SF1 and SF3 data, More on this later.  Further improvements are coming to ensure the reliability of 2010 ACS social and economic data.  Of course, like SF3, ACS is based on samples.  SF1 will continue to be the "nose count of the entire population," but budget exigencies will restrict it to basic age, sex, race, ethnic, and housing information.
   ** The 5% Public Use Microdata Sample (PUMS) of Census 2000,  released August, 2003.  The PUMS files, unlike other census data, are based on individuals, not geographic areas.  With PUMS one can choose a universe of individuals that meet certain characteristics  (all people without a high school diploma who earn over $100,000 per year, for example).  Then, this particular universe can be described in detail, using the census data in the PUMS file (how many spend more than an hour getting to work, for example).  PUMS files based on American Community Survey intercensal estimates are now being released, including for the first time a multiyear PUMS file (for 2005-2007).
   ** In December, 2008, the Census Bureau's ACS released the first nationwide set of multiyear population estimates, covering 2005-2007.  This is a prelude for the new social and economic data that will come in the next decennial census (2010). Sample sizes are small, but the data for states and congressional districts (and for large counties like The Bronx) are widely considered to be trustworthy, as are large counties and a special geographic unit, the PUMA (Public Use Microdata Area).  With PUMAs we have measurements down to the large neighborhood level (a PUMA may be as small as 100,000 population).
   ** ACS has also been releasing yearly census updates since 2000.  The most recent, covering 2007, were released in December, 2008.  Data is available for the US, all the states, large counties, and  PUMAs.  As will be noted in Section 9, the new multiyear data is reliable down to the PUMA level, while single year data for smaller units such as PUMAs may not be as reliable as one would like (sample sizes can be around 3% for the multiyear files, but well under 1% for single years).  But the figures may help us spot trends.
   ** County-level estimates for 2007 have been released by the Census Bureau's Population Estimates Program. 

All the information mentioned above, and for that matter most of the material in this document, is based on data  available to anyone with the patience to navigate the Census Bureau's internet website, http://www.census.gov.  Very helpful tools have been recently introduced;  you are encouraged to get your own data by carefully following directions at the Census Bureau's American Factfinder website.  Tabs at the top of the American Factfinder window allow you to choose the 2000 Census or ACS or  Population Estimates.

Unless otherwise indicated, the statistics in this study come from the various Census Bureau sources mentioned above.

Now let us proceed to the major sections of this website.
 

SECTION 1.  THE BRONX COMPARED TO ALL OTHER US COUNTIES: UNIQUE FEATURES

In 2000, 48.4% of The Bronx population was Hispanic.  By 2007, ACS estimates put the figure at 51%.    In the eastern half of the United States, only Miami-Dade County, Florida has a higher proportion of Hispanics.  All the other counties with higher Hispanic rates are in the southwest, mostly along the Mexican border.

Over 30% of The Bronx population refused to identify themselves with "traditional" racial categories in the 2000 census.   Instead of the traditional categories, people identify themselves either as "some other race", or as "multiracial."  This figure rose to  36.3% in 2004, and to 39.8% by 2007.   In the 2004 data for all counties,  no other county exceeded 30% in the eastern half of the US.

In 2000, only 14.5% of the Bronx population identified themselves as "Non-Hispanic White."    The ACS 2007 estimate is 12.7%.   In the eastern half of the US, only three counties have a smaller percentage of non-Hispanic Whites: the Menominee Indian reservation in Wisconsin, and two overwhelmingly Black counties: Jefferson in Mississippi, and Macon in Alabama.

In 2000, the Bronx was one of five counties in the US where over 30% of households consisted of families headed by single women (2007: 30.0%).  Three of the five are Indian reservations in South Dakota; the fourth is Holmes County, Mississippi.

<>ACS 2007 figures were made available in August, 2008 for all counties with at least 65,000 population.  They confirm Census 2000 data on Bronx poverty: in 2007, the Bronx County child poverty rate of 38.1% was the highest in the Northeast (north of the Ohio River and east of Illinois).  27.1% of the whole Bronx population was below poverty in 2007, also the highest figure in the Northeast.    According to ACS 2005-2007, Congressional District 16 (entirely in The Bronx, covering most of the South Bronx) was the only congressional district in the US with a poverty rate over 40% (next highest figure was 33%).  And it was the only congressional district in the US where a majority of children were below poverty.   And no other congressional district had so few people "well off" (household incomes five times the poverty rate and higher).

2

BACK to Table of Contents.


SECTION 2.  THE BRONX IN THE NEW YORK METRO AREA: ZIPCODE COMPARISONS

In the New York City area, ZCTA's, or Zipcodes, as we shall usually refer to them, average around 40,000 people.  They are large enough to have meaningful data even for small population groups, but there are enough of them to show neighborhood differences. This section contains dozens of maps and considerable analysis of socio-economic trends.  There are four major subject areas:    1.  Racial-Ethnic Differences
                                                                              2.  The Labor Force
                                                                              3.  Education and Other Social Indicators
                                                                              4.  Poverty
 

For those who want to experiment with hundreds of these variables in an Excel file, go to New York Metro Area Zipcode Statistics for 2000.

Here are three maps that illustrate how this material is presented.  After the three, click on the link to access dozens more.  And don't forget to come back here!


 

CLICK HERE FOR  COMPLETE  SECTION 2:   ZIPCODE MAPS AND ANALYSIS

BACK to Table of Contents.
 

SECTION 3.  THE BRONX THROUGH TIME
 

You may want to begin by looking at a list of key items - 198 of them - taken from decennial censuses 1960-2000.  But then you should see how graphic representation and some comon sense analysis can make the figures more meaningful.  To illustrate the latter point, here are a couple of charts:

To see the list of 198 items, click on  comparative data list.

 CLICK HERE FOR COMPLETE SECTION 3:  THE BRONX THROUGH TIME.

BACK to Table of Contents.


SECTION 4.  INSIDE THE BRONX: ZOOMING IN ON RACIAL AND OTHER CATEGORIES

In contemporary America, "race" is an evolving notion whose only constant is its ambiguity. To see some elaboration on this theme, please click on racial changes and ambiguities.   As part of the elaboration we suggest alternative ways to deal with those choosing more than one racial identification.   Click on Census 2000 Racial Data for NY State, NY City, The Bronx, and Vicinity for details.  For government attempts to "modify" the racial identification of millions of Hispanics, check out this website.

At the heart of this section are depictions of ethnic and racial change during the past decade.  The next two maps illustrate what you will find.  Click on the next hyperlink to see more.


 

 CLICK HERE FOR COMPLETE SECTION 4:  Ethnic and racial change inside The Bronx.

BACK to Table of Contents.
 

SECTION 5.  EXTENDING THE ZOOM TO NEW YORK CITY: RACE AND OTHER CATEGORIES
 

It is useful to compare The Bronx with the much larger city in which it is located.  We see in New York City as a whole much larger areas dominated by Whites and by Blacks, while the Hispanic center of gravity for the whole city (indeed for the whole northeastern United States) is in the single area covering The Bronx and upper Manhattan.

 CLICK HERE FOR COMPLETE SECTION 5:  New York City race and other categories.

BACK to Table of Contents.
 

SECTION 6.  ARE THERE "TYPICAL" BENCHMARKS TO COMPARE WITH BRONX REALITIES?

Here we see perhaps the most dramatic difference between The Bronx and just about everywhere else.  Click on the next hyperlink to put the chart into a larger context.

 CLICK HERE FOR COMPLETE SECTION 6:   BENCHMARKS AND "NORMALCY".

BACK to Table of Contents.
 


SECTION 7.  CHANGING SOCIO-ECONOMIC REALITIES IN THE BRONX

a.  Comparisons with Other Boroughs

   The next two charts show differences between The Bronx and the other boroughs in a multidimensional framework.  You will understand more by clicking on the hyperlink to see the whole section.


  CLICK HERE FOR COMPLETE SECTION 7a:  BRONX AND OTHER BOROUGHS.

BACK to Table of Contents.


   b.  Detailed Analysis:  Relation of Race to Poverty

    Using the detailed information from the Census 2000 Public Use Microdata or PUMS files, we can fine-tune the ethnic categories considered thusfar (non-Hispanic White, non-Hispanic Black, non-Hispanic Asian, and Hispanic), to take into account more subtle Bronx demographic realities.    As observed elsewhere in this document, there are significant differences between native born and foreign born Blacks (the latter overwhelmingly from the West Indies).  It also makes sense to study differences between Puerto Ricans, Dominicans, and the other "foreign" Hispanics.  In fact, with PUMS we can study at least seven distinct racial/ethnic groups in The Bronx,  each large enough to present reliable statistics.  Unfortunately, the PUMS files after 2000, like the ACS data down through 2006,  have much smaller samples than the "5% PUMS" for 2000.   We prefer to wait for more reliable sample sizes before the following material is updated


 

    With the details from Census 2000 PUMS, we can sharpen our focus even more.  For example, we can see patterns of wealth and poverty within each of the seven groups.  We have arbitrarily created four household income categories as follows:

        Category Income, Relat. to Poverty       Med. HH Income (2000) of Category

        Below Poverty        Below Poverty Line                    $ 6,400
        Near Poverty          Pov Line to 2x Pov                       19,000
        Mid-Level               2x to 4x Pov Level                       35,600
        High Income           Over 4x Pov Level                       70,000

In the following chart, note the differences between native and foreign born Blacks.  The foreign born Black pattern seems to be closest to non-Hispanic Whites.  On the other hand, there is no significant difference in patterns among the three Hispanic groups.

    PUMS allows us to dig deeper.  We can control for differences in household structure among the seven categories.  For example, single mother households are not spread evenly among the seven racial/ethnic categories, nor are the elderly.  To minimize these problems, let us limit the data to men, age 30-54.  In the next chart we show how many such men are in each household income category.  Note that in 27 of the 28 columns there are more than 2000 people, probably enough for reliable statistics.  When we compare these middle aged men with households in general, the men show lower poverty rates.  Still, patterns of income distribution are similar to the previous chart.

    Education levels also differ among the seven racial/ethnic categories of middle aged men.  Using PUMS, we can refine the data even further by specifying education level.  Note the unique pattern for non-Hispanic Whites.  Only among them do college graduates significantly exceed those with no high school diploma.

    When middle aged men have college degrees, is it more likely that they will be in high income households?  Yes indeed, the relation between education and high income is similar for each of the ethnic categories.  However, well educated non-Hispanic White men are by far the most likely to be in high income households.  For native and foreign born Blacks as well as Puerto Ricans, a bit more than half of these well educated men are in high income households.  For Dominicans and other Hispanics, high education has not been so lucrative.  Without making a definitive comment on race and wealth, we can note that non-Hispanic White men with high school degrees are much more likely to be rich than Dominican or "other Hispanic" men with college degrees.

    Things are reversed when we consider the percentage of middle aged men below poverty.  Of course, poverty is not the "opposite" of high income.  Indeed, the poverty chart seems a bit more complex than the high income chart.   Middle aged non-Hispanic White men lacking a high school diploma are more likely to be in households below poverty than foreign born Blacks.   Of course, as we have seen, there are few such non-Hispanic White men in the first place.  Education protects everybody from being poor, but it protects non-Hispanic Whites better.

BACK to Table of Contents.
 

   c.  Recent Patterns of Migration into and out of The Bronx

    From 1995 to 2000, Bronx population increases have been due to a high natural increase (births minus deaths) and a net increase of migration from foreign countries.  On the other hand, internal migration (movement within the US) shows a larger outflow than inflow.  Migration to and from Puerto Rico, a special case in census statistics, must also be considered because it is so significant for The Bronx.  Outflow back to  Puerto Rico was far greater for The Bronx than for any other US county.  On the other hand, inflow from Puerto Rico was even larger, and was exceeded only by Orange County, Florida (Orlando).    For all Census 2000 data on the latest 5 year migration patterns, CLICK HERE.

The following chart shows internal migration patterns between The Bronx and every county where inflow+outflow exceeded 4000 people between 1995 and 2000.  Note the net increase in migration to The Bronx from core areas like Manhattan and Brooklyn.  On the other hand the migration away from The Bronx to the suburbs greatly exceeded the movement in the opposite direction.  The large migration to The Bronx from little Allegany County, in the southwest corner of New York, is a mystery not yet explained.   Check the New York Times article that gave me my "15 minutes of fame."   For county-to-county migration data, CLICK HERE.

BACK to Table of Contents.
 

SECTION 8.  THE WELL OFF" BRONX

In Section 1, we noted a number of areas where The Bronx was among the poorest counties in the country.  But that is not the only reality here.   There have always been many well off, even affluent people in The Bronx, and this group is expanding, thanks to increasing affluence among  Blacks and Hispanics.  Using Census 2000 PUMS data, we have calculated that there are approximately 417,000 Bronx residents with household incomes above $50,000 ($10,000 above the national median household income in 2000).  This figure is larger than the total population of many important American cities, such as Atlanta, Minneapolis, Miami or St. Louis.

How different are the well off  from the rest of Bronx residents?   With the Census 2000 PUMS file we can select all 417,000 well off Bronx residents and compare many of their social characteristics with various geographic benchmarks.  Here are two examples.

CLICK HERE FOR COMPLETE SECTION 8:  "Well Off" Bronx.

BACK to Table of Contents.
 

SECTION 9.    LATEST ONE YEAR AND MULTI-YEAR UPDATES THROUGH 2008

NEW LOCAL DATA:  SINGLE YEAR AND MULTI-YEAR

<> In intercensal years, the Census Bureau releases  updates for congressional districts, counties, states, and the entire United States.  Since 2005, figures for a smaller unit, the PUMA, have also been updated.  The latest estimates cover a three year period, 2005-2007, and single-year data for 2007.   Both data sources are easily available through Fact Finder on the Census Bureau's website.   The ACS figures were used in the 2007 Small Area Income and Poverty Estimates (SAIPE).  Finally,  the experimental program that includes The Bronx among 36 US counties came out with five-year estimates (2001-2005).   With such a variety of sources, we can check the accuracy of one dataset against the others.  However, each has its limitations.  Budget problems have forced cutbacks:  the Bronx test estimates for 2001-2005 are based on interviews from 6% of households, about half of the sample for the 2000 SF3 file, and so small that tract-level figures might lack requisite reliability. (See the evaluation by Joseph Salvo et al).   Figures are lacking for three of the ten Bronx PUMAs, as well as for six census tracts and dozens of block groups (see the technical evaluation in the Census website).      In the 2007 single-year ACS file for The Bronx, the sample is 14,890, around 1.1% of the Bronx population.  Much information may be distorted with such a small sample.  For example, this file estimates the Bronx population of Mexican ancestry at 68,061, but with a margin of error of 10,160. 

For population sub-groups there may be reliability problems.  For example, from 2005 to 2006 in the single year files, the poverty rate for PUMA 3710 in the south Bronx went from 52.2% to 43.2% (margin of error +/-3.9%).  The Census Bureau warns not to compare 2005 and 2006 PUMA poverty data because in 2006, including group quarters population (the huge Rikers Island prison is in PUMA 3710) may increase the poverty population.  So the big decrease is certainly not credible.  More likely, endemic problems and new procedures pose special problems to the Census Bureau:  the ACS requires a different staff and different timelines from the classic decennial census;  group quarters data, first added to ACS in the figures for 2006, may be difficult to incorporate (group quarters population is omitted from PUMA 3702 in the north Bronx);  ACS total population estimates within large counties are required to add up to the figure from Population Estimates, which use a different  methodology;  budget difficulties may require undesirable compromises, especially for an ongoing program like ACS.   All this can make some year-to-year PUMA data unavailable or unreliable, but it may still be useful to comare the figures we have  for each recent year with the more reliable multiyear data now coming out.  

INTRODUCING THE PUMA

PUMAs, originally the smallest unit for the Public Use Microdata Sample (thus the name), now will become the key local area for the most recent census figures, at least until the 2010 census is released.  PUMAs were designed with a minimum population of 100,000, and usually do not exceed 200,000,.  No PUMA crosses state lines, and for larger counties they do not cross county lines either (one PUMA may include two or more smaller counties so it can attain the minimum 100,000 population).  Where possible, PUMAs reflect local political and social realities.  So in The Bronx and the rest of New York City they often resemble Community Districts - but the boundaries are not exactly the same.  The Bronx is divided into ten PUMAs, and the diversity of the borough is easier to measure.

We can also select the PUMAs that define "the New York Metropolitan Area" as it exists in most people's minds, covering northern New Jersey and southwestern Connecticut as well as New York City, Long Island, and the northern suburbs.   This is the area where the "home teams" are the Yankees and the Mets, not the Red Sox or the Phillies.  Ivy Leaguers would say it extends from Princeton through Vassar up to Yale.  This larger area contains over 20 million people, and is divided into 153 PUMAs.   The second map below shows detailed PUMA boundaries for the entire NY Metro area.






The structure of PUMAs enables us to be more flexible about geographic comparisons.  Our focus remains on The Bronx and its ten PUMAs.  But we can make comparisons with other geographic units, including all 2071 PUMAs in the United States.  When we do so, we can see much finer distinctions than with Congressional Districts:  there are more than 5 times as many PUMAs, and their boundaries are usually drawn to reflect local neighborhood realities - definitely not the case with Congressional Districts! 

The 2005-2007 multiyear files released in spring, 2009 allow us to make interesting detailed comparisons of Bronx neighborhoods with benchmark figures from the wider world.  In the dozen charts that follow, each of the ten Bronx PUMAs is compared with the US as a whole, as well as with New York State, New York City, Bronx County, and all the other Bronx PUMAs.  From the Bronx up to the USA, each geographical unit is a part of  the preceding unit.  All ten PUMAs aggregated together constitute the entire Bronx. 

<>The first four charts show ethnic and linguistic diversity.   Hispanics are so predominant in the western half of the Bronx that they give the entire borough a Hispanic majority.  But in the eastern half there are large enclaves dominated by Blacks or non-Hispanic Whites.  These three ethnic designations do not imply three homogeneous ethnicities.  There are vast differences between native born and foreign born Blacks, and within Hispanic areas,  Puerto Ricans and Dominicans seem to go their separate ways.  In the poorest PUMAs the Spanish-speaking population is much more linguistically isolated - that is, no adult in a family speaks English  "very well".  And finally, as noted elsewhere, Bronx Hispanics simply do not accept the traditional racial categories used in the US.






The next chart shows the impact of recent immigrants (most from Mexico or the Dominican Republic) on the poor neighborhoods of the southwest Bronx.  Note PUMA 3703 (an area with many middle class homeowners) seems unusual among Bronx PUMAs, but it is the only one that resembles the national "benchmark."

In the next two charts we see the family structures kids live in, and the economic consequences of family structures.  The great majority of kids throughout the US live with two parents, and this is also true for kids in the more affluent and more White parts of the Bronx.  But in most of the Bronx, single mother families clearly predominate.  And, as the second chart reveals, the  poverty rate is dramatically higher for single mother families everywhere.  Note also that the mainly Black middle class PUMA 3702 and the mainly White middle class PUMA 3703 both have lower poverty rates than the national "benchmark."  For 3703, the poverty rate for kids in single mother families is around 25% - six times the poverty rate for kids in married couple families in the same PUMA.  But the 25% figure is lower than for kids in married couple families in the four poorest PUMAs.


Back in 2005,  PUMA 3710 in the extreme south Bronx had an unusual concentration of poverty when compared with the 2070 other PUMAs in the United States.  With 52.2% of the population below the poverty level, it was statistically tied with PUMA 6701 (in Hidalgo County Texas, on the lower Rio Grande) as the poorest PUMA in the US.  The rate of the Hidalgo PUMA was 53.7%, but the margin of error for both was  +/-5.9%.     Rates of extreme poverty (incomes half the poverty rate or lower) were even more striking in 2005.  In PUMA 3710, 34.2% of the population was in extreme poverty.  The  next highest  PUMA was 29.1% (in Hidalgo County, Texas)  and the third highest was 24.6% (in inner city Milwaukee).    Poverty is also extremely concentrated.  The four South Bronx PUMAs (3710, 3708, 3707, 3705) contain about 230,000 people below the poverty line - within an area of around 12 square miles.  More recent data, particularly the multiyear estimates for 2005-2007, show some improvement in the Bronx.

The next two charts focus on income and poverty from the 2005-2007 multiyear estimates.  Median household income is a powerful indicator of most people's economic situation, and the  simple line  shows how different the poor half of the Bronx is from  the rest of the Bronx (which is quite similar to  national, state, and city  "benchmarks").  As the aggregate of all  ten PUMAs, the Bronx figure by itself does not show the difference between these two parts of the borough.    But we can also show the extremes of wealth and poverty.  As calculated by the Census  Bureau all figures based on the poverty line are on a sliding scale depending on family size, number of children, and numer of seniors.  Very roughly, for a family of two adults and two kids in 2007, the poverty line is around $21,000 and the "well off" figure (4X the poverty line) starts around $84,000. 


Finally,  three charts show the differences in labor force participation and its effect on poverty.  Following Census Bureau practice, figures for men and women are considered separately.   The first two charts include all those who work, with the bars indicating what proportion  work full-time and what proportion part-time.   The lines show how much the two classes earn, and dramatically highlight the constant gap between  full-timers and part-timers.  Again, for full-time earnings, the more middle class Bronx PUMAs look similar to the "benchmarks", but full-time male workers earn much less in PUMAs 3705 to 3710.  For the female workforce, note the unusually large proportion of full-time workers in middle class Black 3702 and middle class White 3703; but for all their work, the women in 3702 have significantly lower earnings.  The last chart measures the men and women who are NOT in the labor force, for whatever reason.  We have limited the age for both men and women to at least 25 (not counting most students and many younger mothers) and not over 54 (eliminating most retirees).  One would expect few of these people to be outside the labor force, and this is basically true for most men, except in PUMAs 3705 and 3710 (Riker's Island prison in 3710 probably skews this figure).   And  women  are more likely to be out of the labor force in those areas where they must raise the kids. 





We must be cautious with the most recent Census Bureau estimates.  Still, the most recent figures show a number of trends since the 2000 census.  Comparing 2000 data with the multiyear (2005-2007) file and with ACS 2007, we see the Bronx population expanding, slowly but surely.  In the series of three charts that follow, we see the overall growth, but as we focus on greater ethnic detail, we see an enormous migration of Dominicans into The Bronx, plus increases of Mexicans, West Indians, and Black Africans.  All these groups are relative newcomers from foreign countries.  The native-born groups who predominated in past years have  not significantly increased (Puerto Ricans) or have significantly decreased in  population (native-born African Americans;  non-Hispanic Whites).


 



The most striking characteristic of The Bronx is the extent of poverty, particularly in the southwestern part of the borough.  Here the latest census figures allow for some cautious optimism.  Poverty rates are lower,  even among the groups that are most vulnerable.  In the following chart we see a significant decline in the poverty rate for single mothers with kids.  The figures also seem to show a decline in the number of families with kids, whether headed by a single mother or by a married couple.    The trends seem to be clear, but it might be prudent to wait a while before constructing new social theories based on this information.  Therefore, the PUMA level material in the next section will not be revised just yet.



<>

The income differences among the NY Metro PUMAs confirm the theme we have seen so many times:  parts of The Bronx suffer from unusually high poverty levels.  The theme repeats itself whether we deal with the overall poverty rate, poverty for babies, or extreme poverty.  In 2007, the majority of kids under 18 was below poverty in four NY Metro PUMAs.
  Three of the four are in the south Bronx.  In fact, PUMA 3710, with a child poverty rate of 58.6%,  is highest in the continental United States (many PUMAs in Puerto Rico are even poorer).  If we go to the opposite extreme,  in most of the South Bronx,  practically nobody can be classified as upper income (even though there is a substantial "well off" population in other parts of the borough).  And only three PUMAs in the New York metropolitan area had an overall poverty rate above 40%.  All are in the south Bronx. 

Median household income figures also reveal an interesting contrast in the New York area.  As the map below shows, the area can be viewed as three very different concentric circles.  The highest poverty rates are in the urban core;  most close-in suburbs are among the richest areas, while the remotest suburbs fall back toward the national mean.





The ethnic balance of The Bronx continues to change.   As we saw in the beginning of this section, in The Bronx as a whole the Dominican population continued to migrate to The Bronx,  and there were also significant gains for Mexicans and foreign Blacks.  Both non-Hispanic Whites and African Americans declined.   When we look at each Bronx Puma,  we see an even more complex reality.

 




Detailed PUMA maps for the entire country are available in pdf format from the Census Bureau.  Click here for the website.    Data for 2006 and earlier years, as well as all relevent maps, can be obtained on the excellent IPUMS website.  Click here.

POPULATION CHANGE SINCE 2000 FROM CENSUS BUREAU'S POPULATION ESTIMATES PROGRAM

In 2008, according to the Census Bureau's Population Estimates, the Bronx population was 1,391,903,  an increase of 4.4% from the official 2000 figure.  These estimates are used within the Census Bureau, even for programs like ACS that use different statistical procedures.  But they can be challenged by local officials.   Click here for an account of New York City's successful challenge of initial census estimates. 

Racial-ethnic estimates for counties have now been released for each year from 2000 through 2008.   The Bronx can be compared with its neighbors and near-neighbors in NYC (Manhattan, Queens, Brooklyn) and in the suburbs (Nassau and Westchester, as well as Bergen County, NJ, across the Hudson).  The overall population of The Bronx seems to be rising faster than its suburban neighbors.  As the next chart shows, it was virtually tied with Nassau County in 2000, but has steadily been gaining on Nassau, and by 2008 it it ahead enough to be seen on the chart.  The Bronx rate of increase is around the same rate as Brooklyn and Queens, but slower than Manhattan.  But, of course, looking at individual ethnic groups conveys a much more complex reality.



The Bronx non-Hispanic White population continues to decline, but this is also the case in more affluent suburbs like Nassau and Westchester, as well as in Queens.  Surprisingly, in Brooklyn and especially in Manhattan, non-Hispanic Whites are increasing.  In fact, in all the other counties they are the largest of the four groups displayed here - but not in The Bronx.



The non-Hispanic Black population seems to be declining in Brooklyn and Manhattan, the two NYC counties where Whites are increasing.  This may relate to the gentrification of parts of the city.  In The Bronx we do not yet see the phenomenon of "Black flight and White return" to the same extent as Brooklyn and Manhattan.  In fact, the proportion of Blacks in The Bronx (almost 31%) is higher than any of these counties except Brooklyn (33%).  If current trends continue, The Bronx may soon surpass Brooklyn in pecentage of Blacks.  However, the unpredictable economy makes it hard to anticipate future trends.


If The Bronx has a uniquely small proportion of non-Hispanic Whites, Queens has a uniquely large population of Asians, from many parts of that continent.  In the other neighboring counties, Asians are expanding, but from a rather small base.  The rate of expansion seems least in The Bronx, where Asians are less than 3.5% of the population (over 20% in Queens).


There are more Hispanics in The Bronx than any other county portrayed here; in fact, more than in any other county this side of Miami or Chicago.  Not only is the population high, it is expanding vigorously, as the chart below indicates.  Perhaps more interesting is the fact that Manhattan, for many years the center of Dominican life in the US, now has a declining Hispanic population.  Hispanics are now 51.7% of The Bronx population.  Among the counties portrayed here, Queens is next (26.7%), followed by Manhattan (24.5%).




  Elsewhere in the US, certain counties have seen dramatic differences since 2000.  Population decline in the New Orleans area after Katrina was, we fervently hope, a unique disaster, but stagnation or even decline seems to have affected many central cities as well as rural mid-America, whereas some suburban counties increased significantly.  The Bronx and the other counties of New York City appear healthier than many other core urban areas.  In the following map we can see regional differences in rate of change over the past six years.
 


<>

The 2006 county estimates are based on four components of demographic change:
        1. "natural increase" (births minus deaths).
        2. "net internal migration" (people moving into the county from elsewhere in the US, minus people moving out of the county to another US address).
        3.  "net international migration" (people from a foreign country or Puerto Rico moving into the county, minus people moving from the county to such areas).
        4.  There is also a "residual" element, which will be considered later.
 

1.  Natural Increase
The Bronx birth rate is the highest in the Northeast, while its death rate is relatively low.  So, in 2006, its rate of "natural increase" remained among the highest in the Eastern US, as the following map shows.


<> 

2.  Net Internal Migration
Like many other urban areas, The Bronx lost a substantial population to internal outmigration.  But the internal outmigration rate was much higher in Brooklyn and Queens.  Non-Hispanic Whites are more likely to move to suburbs, Florida, or other American destinations than "minorities." Most of the "white flight" from The Bronx occurred before 2000; now it seems to be happening in Brooklyn and Queens.  But internal outmigration is also occuring in small rural areas in the Great Plains, in Appalachia, and even along the California coast.  The national map shows how widespread it is.


 

3.  Net International Migration
International migration keeps the New York metropolitan area from suffering a disastrous loss in population.  But even in Brooklyn and Queens, with extremely high rates of international inmigration, it is not high enough to counterbalance the outflow to other parts of the US.  As the national map shows, the northeast corridor attracts a heavy inflow of foreigners, as does the southwest, south Florida, North Carolina, and other scattered areas around the country.  Each has its own type of immigrant.  It is Dominicans who make the Bronx shine deep red on this map.


 

4.  The "Residual" Category
Census Bureau statisticians also use a "residual" category in calculating their population estimates.  Residuals are included in the "entire dataset" (COEST2005-alldata) that can be downloaded from the Census Bureau's Population Estimates website.  Residuals are used to raise or lower figures for counties so that they add up to the national population figure that is already determined.  They may also be manipulated to reflect revised estimates, even for past years.  Thus, New York City produced credible data and persuaded the Census Bureau that recent estimates for Brooklyn and Queens were too low.  Residuals for past years were revised by the Census Bureau to reflect such corrections.

Residual population figures usually amount to no more than a few hundred added or subtracted, even for the largest counties.  But a few ( 18 large counties, including Brooklyn, Manhattan, and Queens, and also Baltimore, St. Louis, King County, WA and Fulton County, GA) often have large residuals.  Over the past six years, among the five boroughs of New York City, residual additions or subtractions seem to swing wildly back and forth, with the most recent years seeing the greatest corrections  (probably in response to successful New York City protests of census population estimates).  Here are the residuals for each year from 2000 to 2006, with the Census Bureau's "2005 version" of such figures in red and its "2006 version" in black.
 

                                  2000           2001            2002           2003            2004          2005        2006

Bronx 05                      -79            -527             -449            -459             -525          -318
Bronx 06                      -33           3307            2708            3131            2351         2641         -181  

Brooklyn 05               -290             251                34           9214            25113         -957
Brooklyn 06               -221             235              612         12123            25666      23569        -302 

Manhattan 05             -79           7686            6199            6280              6344          -172
Manhattan 06             -38           9640            8074            8014              2426       13487          10

Queens 05                 -325         10928            8503                76           16600        -1041
Queens 06                 -261             232              691          17716           16365       12584       -264

Staten I. 05                  32            -209             -145             -161              -159             69
Staten I. 06                  35            2209           1754             1633             1991         1910          69

Residuals may be a necessary statistical tool, but from a local perspective, areas like Brooklyn and Queens (and since they furnish almost 60% of New York City's population, the city also) may present problems.  Do statistical techniques reduce the accuracy of demographic estimates?  When they look at the original 2005 figures and compare them with the 2006 revisions for that same year, will laymen (including members of Congress, who pay the bills) understand how and why some residuals have been changed so much?
 

RECENT ACS PROFILES OF THE BRONX

The most recent yearly versions of the American Community Survey present extensive demographic and socio-economic data for each of the states,  counties and cities with populations of 65,000 and up, and, as noted above, PUMAs.   The information is presented on the American Factfinder website in four different ways: Data Profiles, with printable tables for many key items;  Multi-Year Profiles, showing changes each year from 2000 through 2007; Ranking Tables, ranking each county or State or City according to many socio-economic variables; and Detailed Tables, where one can choose from the complete list of ACS variables for one or more geographic areas.  There is also a Download Center where multiple files for all major counties can be obtained.  

The Bronx has basically the same unique qualities today as it did in 2000, as we have noted at the beginning of this study.  Here is what we must emphasize:
                Hispanics are now an absolute majority of the Bronx population.
                In the Bronx, Spanish is the "home language" for more people than English.
                The Bronx is now "Center of Gravity" for Dominicans - at least 220,000 in The Bronx, more than in Manhattan
                With 12.7% non-Hispanic Whites, The Bronx has the third lowest percentage among the 236 counties with at least quarter-million population.  Only Hidalgo and Cameron Counties in Texas are lower.  Among the 70 cities with at least a quarter million population, only Detroit, Santa Ana California, Miami, and Newark have lower percentages of non-Hispanic Whites.
                Among the 473 counties showing this variable, only three had higher child poverty than The Bronx:  Hidalgo and Cameron in Texas, and Orleans Parish, Louisiana..

The Real Significance of Recent Data

In most counties, the most recent population figures are not much different from those of Census 2000.  But (at least for areas like The Bronx) small population changes do not mean stability.   For example, from  2000 to 2006, The Bronx population increased by almost 29,000.  But there was a net gain of  more than 78,000 births over deaths, and a net gain of  85,000 immigrants from abroad, balanced by a net loss of more than 148,000 moving out to other parts of the US.  As seen above, residuals added almost 14,000 to this total.  

Population estimates released in March, 2009 show that net internal migration out of the Bronx has significantly decreased each year since 2005, when the loss was over 33,000.  For 2008 the loss was only 16,000.  During the same period, net international migration held steady, with gains over 10,000 each year.  We must wait for the 2010 census to see if these healthy trends can survive our current economic difficulties.

Each of these larger population shifts has resulted in new challenges for social services and every other aspect of life in the community - indeed, we need constantly to redefine what "the community" is.
 
 

BACK to Table of Contents.

AFTERTHOUGHT:   HOW RELIABLE ARE CENSUS FIGURES?

The many problems of ACS, especially for 2006, have already been noted.  ACS also noted that its 1999 figures for The Bronx are unreliable.  The methodology for estimating race among Hispanics is one of the main problems the Census Bureau has been wrestling with, both for ACS and for the general Census 2000 files.   Indeed, as noted here, this problem seems to puzzle most demography researchers.  When figures are based on estimates from a sample, as is the case with ACS,  once you get one thing wrong, a cascade of difficulties follows.  So as we can see in Factfinder, ACS, in 1999, estimated fewer than 130,000 non-Hispanic Whites in The Bronx;  the 2000 Census SF1 figure is 193,651.

Even more questionable (though perhaps necessary) is the Census Bureau's "modified race data", where everybody who refused to identify with an official "Office of Management and Budget Race" was assigned one of these races.  This was done to meet the needs of certain state and local agencies and for making population estimates and projections, according to the Census Bureau.  Hispanics were most affected, since they were over 90% of those who called themselves "other race."  Miraculously, once the government applied the modifications, the percentage of Hispanics in the US categorized as "White" rose from 47.89% to 92.13%  !  It is instructive to read the Census Bureau's report on modified race data.  Do US government agencies like OMB really understand the changing feelings toward our "traditional" racial categories?  Is there sufficient sensitivity to the self-identification of America's 35 million Hispanics?

You can juxtapose SF1 detailed Hispanic figures with the estimates from ACS.  Go to Factfinder
website  and choose Census 2000 - Summary Tables - Detailed Tables - PCT11. HISPANIC OR LATINO BY SPECIFIC ORIGIN.   Then choose ACS 2000 - Summary Tables - Detailed Tables - PCT006. HISPANIC OR LATINO BY SPECIFIC ORIGIN.  SF1 cannot allocate almost 100,000 Bronx Hispanics to a specific country of origin;  for ACS the figure is only around 26,000.  So ACS shows about 40% more Ecuadorians and 20% more Dominicans than SF1.  ACS also shows 20,000 more Puerto Ricans, but it shows about 15% fewer Mexicans!  Our Census Bureau should provide detailed explanations for these differences.

Despite assurances that the SF1 count is reliable, the Census Bureau has admitted past inaccuracies, some of which dramatically affected The Bronx.

Adjusted figures from the 2000 Accuracy and Coverage Evaluation (A.C.E.) sample survey have been termed "severely flawed" by the Census Bureau and many statisticians.  Estimates for 2000 from Demographic Analysis also seem untrustworthy, particularly with regards to immigrants.  But the Census Bureau's disclaimer  recognizes only that "the adjusted Census 2000 data are...not better than the unadjusted data."  (This disclaimer as well as the A.C.E. adjusted data may be found on a site maintained by UCLA.  Click here.)

If, indeed, adjusted data are no better than the official unadjusted census, the adjustments may still be useful in showing areas where official counts may be questionable.  When we compare adjusted and official figures for tracts in the five New York City boroughs  plus nearby suburbs, we see a pattern of undercounting in poorer urban areas, and most particularly in The Bronx.   This is consistent with the undercount for 1990, as described and mapped below.


 

Focusing on the four major ethnic groups, we see a familiar pattern: non-Hispanic Whites tend to be overcounted, while Hispanics and non-Hispanic Blacks are significantly undercounted.  We cannot accept these maps as valid adjustments of 2000 census data, but they suggest that the Census Bureau must make even stronger efforts to track down minorities, particularly those who are recent immigrants.


 


 


 

BACK to Table of Contents.

The Legacy of 1990:  The Post Enumeration Adjustments

After the official 1990 census results were in place, a Post Enumeration Survey was conducted to include people missed in the original count.  In The Bronx, an additional 62,000 people were "found," a gain of over 5%. Adjusted figures show an undercount of over 6% for Blacks and Hispanics, while the non-Hispanic White population was actually overcounted.  But original 1990 census figures are still the official ones, and were used as the base point for yearly population estimates all the way to 2000.   A map of the under (over) count by area shows that almost all the areas in the "White Bronx" were overcounted, while the South Bronx was significantly undercounted.  A detailed assessment of the problems can be found by clicking on the link.
 


In the next few years (if Congress comes through with the funding), we should be able to update census data every three years, thanks to an expanded American Community Survey.  Then we can  better guess where The Bronx is headed in coming decades.    In the meantime, it may be interesting to look back on The Bronx in 1990.  Many of the socio-economic realities we examined a decade ago remain today; other trends may have been altered.  Readers are encouraged to compare this study of Bronx 2000 with the 1990 study and make their own judgments.  Click on the following link to continue.

BACK TO THE BRONX FROM THE VIEWPOINT OF THE 1990 CENSUS

                                                    - - - - - - - - - - - - - -

<>Since this website uses only census data as the raw material to examine The Bronx, the view presented here is clearly incomplete.  Here are some links to sites that will give further information about The Bronx.
   
    Queens College Social Explorer
    Lewis Mumford Center, SUNY Albany

    Bronx on the Web - N.Y. Public Library
    Bronx Historical Society
    Bronx D.A.'s Office
    Gotham Gazette (NYC Citizens Union)
    NYCHANIS  (interactive databases and maps with census, school, and crime data down to tract and precinct levels)

 We want your comments and suggestions.  Please contact The Bronx Data Center through EMAIL:  william.bosworth@lehman.cuny.edu
BACK to the top.

Last Reviewed  May 16, 2009.