Moving … County-to-County

Bookmark and Share

Written by: William Koerber and Melanie Rapino

Over the course of a year, how many Americans move to a new county? About 6 percent of the population age 1 and older, according to American Community Survey data gathered from 2007 to 2011.

The latest version of the county-to-county migration products are now available, including the updated Census Flows Mapper.  In 2011, there were 3,221 counties or county-equivalents (e.g., Louisiana parishes, Alaska boroughs, Virginia independent cities) in the U.S. That results in 10,371,620 (3,221 x 3,220) combinations of moves people could make between counties. 

 During the data collection period, almost 17 million people moved to a new county over the course of a year.  However, not all counties saw moves between them. In fact, most did not. The survey captured migration between 262,196 county pairs, averaging 64 movers between counties. The largest flows were from Los Angeles County to San Bernardino County, Calif. (41,764) and from Los Angeles to Orange County, Calif. (40,794).  (The estimates are not statistically different from each other.)

Our tool, the Census Flows Mapper, based on county-to-county flows, allows us not only to see where people are moving to and from, but also the net gain or loss for a county. 

Census Flows MapperThese migration products include the full county-to-county migration flow files, which are based on the current residence of respondents and the residence one year prior.  The estimates represent the number of movers between counties in a 12-month interval over the five-year period, similar to annual migration patterns averaged over five years.  Multiple years of American Community Survey data are used in order to get a more robust dataset. 

Because the American Community Survey is a sample of the population, there is a potential variability in the estimates.  We statistically measure this by using the coefficient of variation (CV), the ratio of the standard error of the estimate to the estimate itself – a measure of relative variability between estimates. 

The table below shows the distribution of county pairs by the number of movers and their associated CVs.  A CV of 0.61 or over means that there is a high potential of variability for those estimates.  For the most part, the estimates with high potential variability or high CVs are smaller flows of 100 or fewer people and have less impact on migration trends for a given county, especially one with a larger overall population. 

The inclusion of all flows, even the smaller ones with high CVs, allow users to collapse counties into larger areas (e.g., metropolitan areas, states, remainder of U.S.) and thus customize their own geographies or flows in addition to other potential research uses. 


Number of Movers






Over 10,000

























0.61 and over






Besides the basic county-to-county tables, the package also includes tables by selected characteristics (educational attainment, individual income and household income), Census Flows Mapper tool for visual representation of the flows, working paper with findings and a PowerPoint tutorial.  Additional tables use minor civil divisions for selected states rather than counties. The minor civil divisions in those states provide a wide range of government services, and in the case of New England, are the primary government unit at the substate level.

Posted in Geography | Tagged , | 1 Comment

Beyond a Bachelor’s Degree: Big Gains for Graduate School Attainment

Bookmark and Share

Written by: Jessica Davis

The number of people with some graduate education now exceeds the number who have just a bachelor’s degree –36 million to 30 million, one of the interesting findings in the newly released 2013 educational attainment table package.

The Great Recession, which started in December 2007, saw the economy and job growth decline. Adults without job prospects will sometimes stay in school to further their education but also with the anticipation that when they finish, employment opportunities will be greater.

A recent Census Bureau release showed a  4.5 million student increase in college enrollment since 2001.  As a consequence  of the increase in enrollment,  educational attainment has also increased in recent years. In 2013, 65.6 million adults age 25 and over had a bachelor’s degree or higher (31.7 percent), up from 50 million in 2003 (27.2 percent).

Population 25 and over who have a bachelor's degree or higher

An interesting aspect of the increase in education has been the rising tendency of people with a bachelor’s degree to return to school and get more education. The number of adults who have completed some graduate school increased 24 percent from 2008 to 2013, from 29 million to 36 million. That is to say, 7 million more people have experienced at least some graduate school education than was the case five years ago. As you can see, going to graduate school for additional education beyond a college degree has become the majority experience for those with a bachelor’s degree.

It is too early to tell how many people will end up with graduate degrees out of the total who have continued their education beyond college. However, a large portion have already completed a graduate degree. The number of adults completing a master’s degree grew by 18 percent from 2008 to 2013, and the number completing a professional or doctoral degree grew by 20 percent. By comparison, the number with a bachelor’s degree who haven’t pursued any further education has shown little growth—only a 2 percent increase over the past five years.

Posted in Education | 2 Comments

Measuring Alternative Educational Credentials

Bookmark and Share

Written by: Robert Kominski and Stephanie Ewert

Social science research has repeatedly shown a link between educational attainment and social and economic outcomes. Although we have insight into how high school diplomas and college degrees shape achievement, what about other avenues of training and skill development? 

 In addition to, or instead of, regular schooling, some people earn educational certificates, professional certifications, or licenses or participate in noncredit courses, on-the-job training, or apprenticeships. Findings released today from the 2012 Survey of Income and Program Participation show that 22 percent of adults held a professional certification or license, and 9 percent held an educational certificate. Therefore, a sizable proportion of the population holds alternative educational credentials independent of traditional degrees. A professional certification or license is awarded by a certification body or licensing agency through an examination process or other predetermined criteria. An educational certificate is typically awarded by an educational institution based on a program of study. 

Until now, federal surveys had not collected data on these alternative education and training mechanisms in a systematic fashion. Our new report, Measuring Alternative Educational Credentials, not only provides the first national estimates of professional certifications, licenses, and educational certificates but also explores their labor market value. 

So, do these alternative credentials pay off in the labor market? Our findings suggest that they do, particularly for those with low levels of educational attainment.  Full-time workers earned more with an alternative credential than without. The median monthly earnings for someone with a professional certification or license only was $4,167 compared with $3,433 for someone with an educational certificate only, $3,920 for someone with both types of credentials, and $3,110 for someone without any alternative credential.

The benefit of having an alternative credential is concentrated among people with low levels of regular education. Professional certification or license holders earned more than those without an alternative credential at each level of education below the bachelor’s degree. Among people with less than an associate’s degree, educational certificate holders also earned more than people without an alternative credential.

For more details on the prevalence of alternative credentials and their relationship to labor market outcomes, see our report Measuring Alternative Educational Credentials.

Posted in Education | Tagged | 2 Comments

Using the Survey of Income and Program Participation to Study Poverty Dynamics

Bookmark and Share

Written by: Ashley Edwards

Some people may see poverty as a continuous state. In fact, it can be temporary. Many people are used to understanding the issue of poverty through annual statistics that provide a snapshot of how many individuals have annual income lower than their poverty threshold. However, estimates of how many people experience short spells of poverty, how many people enter and exit poverty and how many stay in poverty for many years can provide important additional insights.

Longitudinal surveys such as the Survey of Income and Program Participation (SIPP) follow respondents over an extended period and collect data at regular intervals. Using data published from the SIPP in the report Dynamics of Economic Well-Being: Poverty, 2009 to 2011, we find that poverty is not a permanent condition. A small fraction of individuals experience extended, or chronic poverty, while a larger percentage of people experience short poverty spells that are not captured in an annual measure.

Because the SIPP allows us to calculate poverty over longer time periods, and track how people move into and exit poverty, we can see how how poverty experiences vary by demographic characteristics in ways that we are unable to capture when looking at annual measures.

For example, annual measures of poverty show that individuals 65 and older have lower poverty rates than children and nonelderly adults. However, when we use the longitudinal nature of the SIPP to look at how individuals enter and exit poverty, we find that although the elderly were less likely than other age groups to enter poverty over the course of 2009 to 2011, it was difficult for them to exit once they entered a poverty spell.

Over the course of 2009 to 2011, elderly individuals were less likely to exit poverty than adults 18 to 64, and their exit rates were not statistically different from the exit rate for children. The median length of poverty spells was also longer for individuals 65 and older, 8.3 months, compared to children (7.0 months) and adults 18 to 64 (6.3 months).


Median Poverty Spell Length by Age, 2009 to 2011

Note: Excludes spells underway in January 2009. Due to changes in the estimation of
survival rates, estimates of median spell length presented in this report are not
comparable with estimates of median spell length reported in previous P70 Dynamics
of Economic Well-Being series.
Source: U.S. Census Bureau, Survey of Income and Program Participation, 2008 Panel.
For information on confidentiality protection and sampling and nonsampling error,

Similarly, when we look at annual poverty rates by race and ethnicity, we find that the rates for Hispanics and blacks were not statistically different. However, this annual measure masks differences in how these groups experience poverty. Over the period from 2009 to 2011, Hispanics were more likely than blacks to enter poverty (10.7 and 8.9 percent respectively) but also more likely than blacks to exit poverty (35.8 and 22.7 percent respectively). As a consequence, the median poverty spell duration for Hispanics was 6.5 months compared to 8.5 months for blacks.

Median Poverty Spell Rate by Race, 2009 to 2011

Note: Excludes spells underway in January 2009. Due to changes in the estimation of survival rates, estimates of median spell length presented in this report are not comparable with estimates of median spell length reported in previous P70 Dynamicsof Economic Well-Being series.
Note: Federal surveys, including the SIPP 2008 Panel, give respondents the option of reporting more than one race. These data can be shown in two ways: (1) as mutually exclusive from other race groups, which may be denoted by “alone” or (2) not mutually exclusive with other race groups, denoted by “alone or in combination with other race groups.” This figure shows race using the first method. Because Hispanics may be of any race, data for Hispanics are not mutually exclusive with race.
Source: U.S. Census Bureau, Survey of Income and Program Participation, 2008 Panel. For information on confidentiality protection and sampling and nonsampling error, see


While the elderly are sometimes overlooked as a vulnerable population due to their declining annual poverty rates, we find that once the elderly enter poverty, they are at a higher risk of becoming stuck in poverty. Similarly, although it may seem from annual poverty rates that Hispanics and Blacks are experiencing similar economic conditions, we find that Hispanics are more frequently churning into and out of poverty, while Blacks are less likely to exit a poverty spell once they’ve entered one.

For additional poverty data from the 2008 SIPP Panel, please see the following link and “Dynamics of Economic Well-Being: Poverty, 2009-2011.”

Posted in Poverty, Uncategorized | Leave a comment

OnTheMap Upgrades – New Business Firm-Level Data and Improved PDF Reports

Bookmark and Share

Written by: Matthew Graham

The Census Bureau has updated the popular OnTheMap tool. Now you can find firm-level statistics, such as firm age and firm size. A product of the Longitudinal Employer-Household Dynamics (LEHD) program, these statistics allow policymakers, academics, and economic development and transportation analysts to measure the economic impact of young/old firms or small/large firms in relation to commuting patterns and worker characteristics.

onthemapOnTheMap is a Web-based mapping and reporting application that allows creating, viewing, printing, and downloading workforce related maps, profiles and underlying data. This innovative tool provides an interactive map viewer that displays employment and residential distributions by the geography you select, available to the census block level. Companion reports provide statistics on employment counts by worker age, earnings, industry, race, ethnicity, educational attainment and sex.

In addition to this new labor force information, you will also see changes to the PDF report output. This complete redesign for each analysis type provides a publication-quality document.  Map, chart, and table elements now appear as full-size, high-quality graphics that are organized in a consistent layout.

What is available in OnTheMap?

  • Comprehensive Coverage – Census block-level coverage for home and work areas in 50 partner states/territories with consecutive years of data from 2002 to 2011. Available characteristics include worker age, earnings, industry sector, worker race, worker ethnicity, worker educational attainment, worker sex, firm age and firm size.
  • Powerful Analysis Tools – Six different analysis options are available for you to analyze, compare, and summarize a vast amount of labor force data for user-defined or Census-standard areas. Each set of results are presented through interactive maps, charts, and reports.
  • Versatile Outputs – Results can be exported to a report (PDF, HTML, XLS), to a map (KML or Shapefile), or to other forms (PNG images or composite PDF). You can also save your analysis settings for future use in OnTheMap.

If you would like to learn more, visit the Longitudinal Employer-Household Dynamics (LEHD) home page or OnTheMap Help pages.

Posted in Uncategorized | 3 Comments