University of Texas at Austin Division of Statistics + Scientific Computation Statistics in Action...

University of Texas at AustinDivision of Statistics + Scientific

ComputationStatistics in Action Series

February 6, 2013Austin, TX

Applied Demography: Some Texas Examples

2

Demography – the study of the size, distribution, and composition of populations; the processes determining these – namely, fertility, mortality, and migration; and the determinants and consequences of all of the above.

~ Bogue, 1968; Murdock & Ellis, 1991

4

The Population Research Institute,

2.1 Kids: Stable Population

http://youtu.be/zBS6f-JVvTY

The Office of the State Demographer disseminates demographic and related socioeconomic data to the State of Texas and the general public. The State Demographer’s Office monitors demographic and socioeconomic changes in the State in order to better inform the executive and legislative branches of Texas government. Special emphasis is placed on data that may be useful to policy makers in dealing with issues regarding the demand for state services.

Mission

• Data Portal, Publications, and Reports http://txsdc.utsa.edu/

• Population Estimates and Projections Program http://txsdc.utsa.edu/Data/TPEPP/Estimates/Index.aspx

• Resource Witness at Legislative Hearings• Public Presentations http://osd.state.tx.us

• Data Requests• Custom Research Projects• Annual Conference for Data Users and Applied Demography

Conference

Meeting the Mission

7

Population Projections, 2010

20000000

25000000

30000000

35000000

40000000

45000000

50000000

55000000

60000000

Projected Population in Texas, 2010 to 2050

No Migra-tion1/2 2000 to 20102000 to 2010

1,000,000

3,500,000

6,000,000

8,500,000

11,000,000

13,500,000

16,000,000

18,500,000

21,000,000

23,500,000

Projected Population for Texas by Race/Ethnicity, 2010 to 2050, 0.5 Migration Scenario

Anglo

Black

Hispanic

Other

Source: Texas State Data Center Population Estimates and Projections Program, 2010 Projections

• Data Portal, Publications, and Reports http://txsdc.utsa.edu/

• Population Estimates and Projections Program http://txsdc.utsa.edu/Data/TPEPP/Estimates/Index.aspx

• Resource Witness at Legislative Hearings• Public Presentations http://osd.state.tx.us

• Data Requests• Annual Conference for Data Users and Applied Demography

Conference• Custom Research Projects

Meeting the Mission

9

Three Examples of Applying Demographic Data and Methods to Government & Business Planning

10

Projecting the educational attainment of the Texas labor force

11

Background

• Texas educational attainment ranked the lowest among the 51 states– 51st in high school graduation– 37th in college graduation

Source: U.S. Census Bureau. 2006-2010 American Community Survey PUMS data

12

BackgroundIndividuals with some higher education are more likely to be employed; trend suggests this gap may be widening.

Source: U.S. Census Bureau. 2001 to 2010 American Community Survey 1-Year PUMS data

2001 2002 2003 2004 2005 2006 2007 2008 2009 201050.0%

55.0%

60.0%

65.0%

70.0%

75.0%

80.0%

85.0%

90.0%

f(x) = − 0.000747696969696972 x + 0.570273333333333R² = 0.0223370054209473

f(x) = − 0.00247781818181818 x + 0.748636R² = 0.441782386010636

f(x) = 0.00167006060606061 x + 0.804954666666667R² = 0.385142289993757

Percentage Employed By Educational Attainment, ACS 1-year PUMS

Less than High SchoolLinear (Less than High School)High School Graduates or HigherLinear (High School Graduates or Higher)Bachelor or Higher

Perc

ent

13

BackgroundEducational attainment has improved in the last 10 years.

Source: U.S. Census Bureau. 2001 to 2010 American Community Survey 1-Year PUMS data

2001 2002 2003 2004 2005 2006 2007 2008 2009 20100%

5%

10%

15%

20%

25%

30%

Educational Attainment Level for Persons Age 18 -64 ACS PUMS 5-year, 2006-2010

< HiHi-EqSome ColAs DegreeBach DegreeMaster DegreeProf DegreeDoc DegreePe

rcen

t

14

Background

< HiHi-E

q

Some Col

As Degree

Bach Degree

Master D

egree

Prof Degree

Doc Degree

0%

5%

10%

15%

20%

25%

30%

Educational Attainment Level for Persons Age 18 -64

ACS PUMS 5-year, 2006-2010

Male Female

< HiHi-E

q

Some Col

As Degree

Bach Degree

Master D

egree

Prof Degree

Doc Degree

0%

10%

20%

30%

40%

50%

60%

70%

Educational Attainment Level for Persons Age 18 -64

ACS PUMS 5-year, 2006-2010

White

Black

US-born Hispanics

Foreign-born Hispanics

Other

Educational attainment improvements seen among most subgroups, but demographic differences are still pronounced.

15

Background

< Hi Hi-Eq Some Col As Degree Bach Degree

Master Degree

Prof Degree Doc Degree0%

5%

10%

15%

20%

25%

30%

35%

40%

Educational Attainment Level for Persons Age 18 -64 ACS PUMS 5-year, 2006-2010

18-24 25-34 35-64

Educational attainment improvements seen among most subgroups, but demographic differences are still pronounced.

16

Data & Methods

• Data sources:– American Community Survey PUMS data 2001 to 2010– Texas State Data Center population projections data– Texas Workforce Commission occupation-education projection

data• Analysis plan

– Multinomial regression model used to predict the rates of each educational attainment category for each demographic sub-group, based on historical trend and demographic characteristics

– The predicted sub-group specific rates are applied to the population projections

17

Findings

• Compare projected educational attainment of the labor force to required labor force for projected job needs

• Identify gaps between demand (by job type) for education and the supply (by the labor force)

• Specific educational attainment levels and population sub-groups can then be targeted to meet these needs– Tailored career paths in high schools, colleges, and vocational

schools– Inform immigration goals for specific skill visas– Evaluate funding needed to meet goals of projected education

levels needed in the labor force

Identifying demographic, socioeconomic, geospatial, and housing unit characteristics that are related to

household energy consumption

• Identified variables related to energy consumption• Data sources:

– Energy provider consumption data– Census data– County appraisal district data

Data & Methods

20

Housing Characteristics

• Year built• Square footage• Type of foundation• Number of rooms• Number of stories• Presence of a pool• Presence of a fireplace

21

Demographic & Socioeconomic Characteristics

• Total population• Total occupied housing units

– Owner occupied– Renter occupied

• Percent Hispanic, non-Hispanic householders• Mean age of householder• Number of persons per household• Family composition• Educational attainment of householder• Households with children under 6• Households with seniors• Median household income

22

Behavioral Characteristics

• Participated in rebate program• Installed a smart thermostat

• Analysis plan– Regression analysis to determine most relevant predictor

variables

• Addressing challenges with data– Differing units of analysis– Geocoding and matching data

Data & Methods

24

Findings

• Identify factors most important to high/low energy consumption

• Identify homes with relatively high energy consumption

• Target homes that could benefit from upgrades, receipt of rebates

• Identify hot spots in the city in need of large scale retrofitting

Estimating the number of unauthorized immigrants in Texas counties

26

Background

• Texas is one of the fastest growing states, with migration making up 45% of growth

• Issue of immigration, especially unauthorized or illegal migration, critical when planning and considering:– Concerns about border security– Assessments of economic impact on receiving communities– Resulting shifts in the social characteristics of communities

• With the exception of California, sub-state level estimates of the undocumented population are not available.

27

Background• Conventionally, estimation of the undocumented population

produced using the residual method (Warren 2011; Passel 2010, 2011) – Estimates of legal foreign born residents are subtracted from

estimates of foreign born

• Most commonly used national and state estimates include Pew Hispanic Center, Dept. of Homeland Security, and R. Warren estimates

• Residual method produces some challenges when attempting to produce estimates at lower geographies due to data unavailability

28

Estimates of the Unauthorized Immigrant Population in Texas

1990 2000 2005 20080

200

400

600

800

1000

1200

1400

1600

1800

450

1,100

1,400 1,400

440

1,127

1,4741,527

1,090

1,360

1,680

Passel Warren DHS

Thou

sand

s

29

Data & Methods

• New county level IRS administrative data allows for reallocation of state estimates of the unauthorized to the county level

• Use regression analysis to account for varying use of ITINs by unauthorized immigrant population

• Expand upon this new estimation technique by employing geographically weighted regression to refine the distribution of unauthorized immigrants across the state

30

Estimates of the Unauthorized Immigrant Population, 2008

31

Estimates of the Unauthorized Immigrant Population, 2008

32

Results

• Higher unauthorized estimates were found in counties characterized by agriculture, urbanicity, high employment, fast Hispanic population growth, and substantial foreign born populations

• These areas include counties in the Dallas-Fort Worth-Arlington, Houston-Baytown-Sugarland, and Austin-Round Rock metropolitan areas, large border counties, and counties in parts of East Texas.

• When examined as a percentage of the county population, Panhandle counties and counties in the Dallas and border areas have higher percentages.

33

Demographics & Destiny

34

http://youtu.be/jbkSRLYSojo

Hans Rosling’s The Joy of Stats BBC Four

35

Contact

Office: (512) 463-8390 or (210) 458-6530E-mail: State.Demographer@osd.state.tx.usWebsite: http://osd.state.tx.us

Office of the State Demographer