Census Report on School Enrollment

8/6/2019 Census Report on School Enrollment

1/20

Population Characteristics

Current

Population

Reports

School Enrollment in the

United States: 2008

This report discusses school enrollment

levels and trends in the population aged

3 and older based on data collected in

2008 by the U.S. Census Bureau in the

American Community Survey (ACS) and

the Current Population Survey (CPS). This

is the second report in a series o reports

using both ACS and CPS data to discuss

school enrollment. The two surveys are

used in this report because o their com-

plimentary strengths in providing data on

enrollment. The ACS is a valuable source

o school enrollment data because o its

large sample size and ability to provide

statistics or small levels o geography.

The CPS data oer a detailed selection o

questions on school enrollment, allowing

or unique analysis.

Highlights o the report are:

EnrollmentinGrades1through12

ell rom 50.0 million in 2000 to

49.3 million in 2008.

In2008,collegeenrollmentwasat

anall-timehighwith18.6million

students enrolled, up 22 percent

from15.3millionin2000.

Two-yearcollegeenrollmentwas

also at an all-time high in 2008, with

5.3 million students enrolled, up

40 percent rom 3.8 million in 2000.

In2008,1millionHispanicstudents

were enrolled in 2-year colleges, up

85 percent rom 540,000 in 2000.

Thenumberofstudentsenrolledin

nursery school in 2008 was not statisti-

cally dierent rom the number o stu-

dentsenrolledin1998;however,there

were changes in the race and

Hispanic origin o the students.

Hispanicstudentsmadeup18percent

o nursery school students, up rom

13percentin1998.

InOctober2008,themajorityof

kindergarten students, 72 percent,

were enrolled in ull-day kindergarten

programs, up rom 28 percent o stu-

dents enrolled in ull-day kindergarten

programsin1978.

THE ACS AND THE CPS:TWO SOURCES OF SCHOOL

ENROLLMENT DATA

The ACS, part o the Census Bureaus

reengineered2010Censusprogram,looks

at a wide range o social, economic, and

housing characteristics or the population

by a multitude o demographic variables.

The ACS is used to provide annual data

on more than 7,000 areas, including all

congressional districts as well as counties,cities,metroareas,andAmericanIndian

and Alaska Native areas with a population

of65,000ormore.In2008,theCensus

Bureau released 3-year estimates rom the

ACS or areas with populations larger than

20,000.Itstartedthereleaseof5-year

estimates or all geographic areas down

tothetractlevelbeginningin2010.The

ACS inormation comes rom a sample o

about3millionaddresses,or1.7percent

ofthenationspopulationeachyear.In

contrast with the CPS, the ACS is admin-istered to the entire resident population,

including those living in institutions and

othergroupquarters.Inthisrespect,data

rom the ACS are highly comparable with

data collected in Census 2000 and earlier

IssuedJune2011

P20-564

By

Jessica W. Davis

and

Kurt Bauman

U.S. Department o CommerceEconomicsandStatisticsAdministration

U.S.CENSUSBUREAU


2/20

2 U.S. Census Bureau

kindergarten, and rst through

twelfthgrade(Table1).

Themajorityofstudentswere

non-Hispanic White (59 percent),

ollowed by Hispanic students

(18percent),Blackstudents

(15percent),andAsianstudents

(5 percent).2 Foreign-born stu-

dents made up 7 percent o allenrolled students and 22 percent o

2 Federal surveys now give respondentsthe option o reporting more than one race.Thereore, two basic ways o dening a racegroup are possible. A group such as Asianmay be dened as those who reported Asianand no other race (the race-alone or single-race concept) or as those who reported Asianregardless o whether they also reportedanother race (the race-alone-or-in-combina-tion concept). This report shows data usingthe rst approach (race alone). This reportwill reer to the White-alone population asWhite, the Black-alone population as Black,the Asian-alone population as Asian, and theWhite-alone-non-Hispanic population as non-Hispanic White. Use o the single-race popula-tion does not imply that it is the preerredmethod o presenting or analyzing data. TheCensus Bureau uses a variety o approaches.Inthisreport,thetermnon-HispanicWhitereers to people who are not Hispanic andwho reported White and no other race. TheCensus Bureau uses non-Hispanic White asthe comparison group or other race groupsand Hispanics. Because Hispanics may be anyrace, data in this report or Hispanics overlapwith data or racial groups.

school enrollment rom the CPS can

be ound on the Census Bureaus

Web site at .

To learn more about the dier-

ences between the ACS and CPS,

see the comparison grid (Appendix

Table A) at the end o this report.

SCHOOL ENROLLMENT

Due to its sample size, the ACS

is used in this section to describe

overall school enrollment in the

UnitedStates.In2008therewere

79.9 million people aged 3 and

older enrolled in school. There

were5.1millionchildrenenrolled

in nursery school, 53.4 million

students enrolled in kindergarten

through twelth grade, and

21.4millionstudentsenrolled

incollegein2008(Figure1).

More emales than males were

enrolled in school but their num-

bers exceeded males only at the

undergraduate and graduate school

levels. More males were enrolled

than emales in nursery school,

decennial censuses.1 The ACS asks

respondents throughout the entire

calendar year whether they were

enrolled in regular school at any

time in the 3 months beore the

interview. The survey also asks

whether each person attended pub-

lic school or private school, and in

what grade or level the person was

enrolled. Tables rom the ACS about

school enrollment can be accessed

through the American FactFinder

on the Census Bureaus Web site at

.

The CPS samples approximately

72,000 housing units each month.

Unlike the ACS, the reerence popu-

lation is the civilian noninstitution-

alized population, so people living

in institutions are not included.While the sample size is not su-

cient to describe small geographic

areas, the CPS is designed to meet

reliability requirements or the

50 states and the District o

Columbia.Estimatesofschool

enrollment rom the CPS are based

on a special supplement, admin-

isteredeachOctobersince1956,

allowing the construction o a time

series o trends or school enroll-

ment. The supplement on school

enrollment asks detailed questions

ofchildrenaged3to14andpeople

aged15andolder.Twentyques -

tions are asked, gathering inorma-

tion on single year o enrollment,

enrollment status and level or

thepreviousyear;whetherthe

respondent goes to school ull-time

orpart-time;whethertheyattend

a2-yearor4-yearinstitution;

whether they are obtaining any

vocationaltraining;andwhatyear

they received their most recentdegree. Tables about students and

1OtherdierencesbetweentheACSandCensus 2000 aect comparisons o schoolenrollment.Oneofthemostimportantisthe reerence time o data collection, whichis the 3 months preceding collection (whichoccurs year-round) in the ACS but is xed tothetimeprecedingApril1inthecensus.Thisdierence especially aects comparisons oenrollment by age.

Nursery

Kindergarten

Elementary(grades 1 through 4)

Elementary(grades 5 through 8)

High school(grades 9 through 12)

College

(undergraduate)

College(graduate or professional)

5.1

4.1

15.9

17.5

16.2

17.2

3.9

Figure 1.

Number of Students Aged 3 and Older Enrolled inSchool by Level: 2008(In millions)

Source: U.S. Census Bureau, 2008 American Community Survey. For information on

sampling and nonsampling error, see .


3/20

U.S. Census Bureau 3

4-year-oldsincreasedfrom1994

to 2008rom 54 percent to

59 percent or 4-year-olds androm 33 percent to 37 percent or

3-year-olds (Figure 2).3

Most o the increase in nursery

school enrollment came beore

3In2008,overhalfofstudentsenrolledin nursery school were 4 years old (59 per-cent) and most o the rest were 3 years old(37 percent). Children aged 5 and older madeup12percentofnurseryschoolstudents.

kindergarten through twelth grade,

high school drop-outs, modal

grade, and college enrollment.

NURSERY SCHOOL

Since the CPS has been collected

or decades, data rom this survey

can be used to examine changes

over time in nursery school

enrollment. The nursery school

enrollment rates o both 3- and

graduate students. Most students

(83 percent) attended public

schools. Among students attendingprivate schools, the largest propor-

tion were enrolled at the under-

graduate college level (29 percent),

ollowed by the nursery school

level(17percent).

The next section will examine

nursery school enrollment and the

ollowing sections will examine

Table1.

School Enrollment by Level of Enrollment and Selected Characteristic: 2008(Population aged 3 and older enrolled in school. Numbers in thousands)

Characteristic

Total

students

Level o school enrollment

Nursery

Kinder-

garten

Elemen-tary

grades

1 to 4

Elemen-tary

grades

5 to 8

High

school

Under-graduate

college

Graduate

schoolTotal 79,855 5,099 4,059 15,923 16,238 17,208 17,452 3,874

SexMale 39,403 2,646 2,081 8,181 8,328 8,849 7,704 1,612Female 40,452 2,453 1,978 7,742 7,910 8,359 9,748 2,262

Age3 to 4 years 4,119 3,886 233 5 years 3,401 1,171 2,117 113 6 to 9 years 15,585 42 1,710 13,687 147 10 to 13 years 15,925 2,124 13,653 147 14 to 17 years 16,391 2,373 13,878 139 18 to 22 years 12,960 19 2,702 10,044 19423 to 24 years 2,126 5 53 1,567 50125 to 29 years 3,221 8 111 2,027 1,07430 to 34 years 1,755 7 68 1,083 598

35 to 64 years 4,161 20 214 2,490 1,43765 years and older 211 6 34 101 70

Race and Hispanic OriginWhite alone 55,922 3,575 2,785 11,021 11,351 11,933 12,469 2,788White alone, non-Hispanic 47,072 2,988 2,211 8,889 9,330 9,985 11,083 2,586

Black alone 11,766 723 571 2,274 2,374 2,762 2,587 475Asian alone 3,797 224 171 651 619 666 1,049 416

Hispanic (any race) 14,440 949 930 3,465 3,295 3,227 2,263 311

NativityNative-born 74,476 5,006 3,956 15,339 15,383 16,048 15,565 3,179Foreign-born 5,378 93 103 584 855 1,160 1,887 695

Type of School

Public 65,990 2,732 3,496 14,136 14,424 15,473 13,428 2,300Private 13,864 2,366 563 1,787 1,814 1,735 4,025 1,574

RegionNortheast 14,141 941 657 2,632 2,780 3,071 3,212 848South 28,904 1,918 1,555 6,017 6,028 6,129 5,994 1,263Midwest 17,621 1,134 881 3,460 3,537 3,818 3,930 861West 19,188 1,105 965 3,815 3,893 4,191 4,317 902

Represents or rounds to zero

Source: US Census Bureau, 2008 American Community Survey For inormation on sampling and nonsampling error, see


4/20


attended nursery school, compared

with 44 percent o those rom

amilies with incomes o less

than $20,000.7

Nursery school enrollment is also

related to the education and labor

orce participation o a childs

mother. Children o mothers whoare college graduates were more

likely to attend nursery school

than children whose mothers did

notnishhighschool(62percent

compared to 33 percent).8 Chil-

dren o mothers in the labor orce

were more likely to attend nursery

school than those whose mothers

were not in the labor orce (54 per-

centcomparedwith41percent).

KINDERGARTEN

As with nursery school, CPS data

are used in this section to describe

kindergarten enrollment because

they can be used to examine

changesovertime.In2008,there

were 4 million students enrolled in

kindergarten, up rom 2.9 million

students3decadesagoin1978.

Kindergarten enrollment has not

only increased in the past 30 years,

but the type o kindergarten enroll-

menthaschanged.In1978,themajorityofstudents(72percent)

were enrolled in part-day kindergar-

tenprograms.InOctober2008,the

majorityofstudents(72percent)

were enrolled in ull-day kindergar-

ten programs (Figure 3).

Black students have historically

been more likely to attend ull-day

kindergarten classes than other stu-

dents.In1981,56percent

o Black kindergartners were

7 Children rom amilies with incomesbetween $20,000 and $29,000 were notstatistically dierent in their enrollment romchildren whose amilies earned less than$20,000 or between $30,000 and $39,000.

8 Children o mothers who are collegegraduates are more likely to attend nurseryschool(62percent)thanthosewhosemoth -ershavesomecollege(56percent),whosemothers are high school graduates (39 per-cent), and whose mothers are not high schoolgraduates (33 percent).

up64percentofnurseryschool

enrollment, ollowed by Black

students(16percent),Hispanic

students(13percent),andAsian

students (4 percent).5

Overhalfofallstudentsenrolledin

nursery school (57 percent) went

to a public nursery school. More

students attended a part-day nurs-

ery school program than a ull-day

nursery school program.6

For some amilies, the cost o

attending nursery school may

prevent them rom enrolling their

children.In2008,56percentof

3- and 4-year-olds rom amilieswith incomes o $40,000 or more

5 There was no statistical dierencebetween the enrollment o Black students innurseryschoolin1998(16percent)andin2008(15percent).Inaddition,therewasnostatistical dierence between enrollment oAsianstudentsinnurseryschoolin1998 (4 percent) and in 2008 (4 percent).

6In2008part-daynurseryschoolatten -dance (52 percent) was higher than ull-daynursery school attendance (48 percent).

1998.Thenumberofstudents

enrolled in nursery school in 2008

was not statistically dierent rom

the number o students enrolled

in1998.Enrollmentof4-year-oldsin1998(60percent)isnotstatisti-

cally dierent rom the percent-

age enrolled in 2008 (59 percent).

Enrollmentof3-year-oldshasalso

remained statistically unchanged

since a decade ago.4

Themajorityofstudentsenrolled

in nursery school in 2008 were

non-Hispanic White (59 percent),

ollowed by Hispanic students

(18percent),Blackstudents

(15percent),andAsianstudents

(4 percent). Ten years ago,

non-Hispanic White students made

4 There was no statistical dierencebetweenenrollmentof3-year-oldsin1998(37 percent) and enrollment in 2008(37 percent). The 90 percent condenceinterval includes zero. The Census Bureaudoes not have sucient statistical evidenceto conclude that the actual change is dierentrom zero.

Percentage of 4-year-olds

0

10

20

30

40

50

60

70

20082007200620052004200320022001200019991998

Percent

Figure 2.

Nursery School Enrollment of 3- and 4-Year-Olds:1998 to 2008

Source: U.S. Census Bureau, Current Population Surveys, October 19982008.For information on sampling and nonsampling error, see.

Percentage of 3-year-olds


5/20


rom 50 million in 2000 to

49.3 million in 2008.10 However,

the decrease in enrollment was

not experienced evenly across

states. Data rom 2000 and 2008

showedthatwhile13statessawan

increaseinenrollmentforGrades1

through12,37statesexperienced

adecrease.Ofthe13stateswith

an apparent increase in enrollment,

only 7 had a statistically signicant

increase.Ofthe37stateswithan

apparent decrease in enrollment,

only16hadastatisticallysigni-

cant decrease (Figure 4).11 Change

in enrollment across states was

almost perectly correlated with

changeintheaged5to17popula-

tion (a correlation o .998).

Enrollmentchangestrackcloselywithchangesintheaged5to17

population. They also correlate

with overall population change in

astate(allages).InMontanaand

Alaska school enrollment declined

by10percentwhileoverallpopu-

lation grew by 3 and 4 percent,

respectively.Incontrast,Nevada

had the highest population growth

rate in the overall population

(24 percent), but even aster

growth in enrollment o 5- to

10In2000,98.2percentof5-to17-year-oldswereenrolledinschooland in2008,96.8percentof5-to17-year-oldswere enrolled in school. There was no statis-tical dierence between the number o 5- to17-year-oldsin2000(53.1million)andthenumberof5-to17-year-oldsin2008 (53.0 million).

11The states whose decrease in rstthrough twelth grade enrollment rom 2000to 2008 was statistically signicant wereCalifornia,Iowa,Louisiana,Maine,Maryland,Massachusetts, Michigan, Minnesota,Montana,NewYork,NorthDakota,Ohio,Pennsylvania,RhodeIsland,Vermont,andWisconsin. The states whose decrease was

not statistically signicant were Alabama,Alaska, Arkansas, Connecticut, the DistrictofColumbia,Hawaii,Illinois,Indiana,Kansas, Kentucky, Mississippi, Missouri,Nebraska, New Hampshire, New Jersey,NewMexico,Oklahoma,Oregon,SouthDakota,Washington,WestVirginia,andWyoming. States with a statistically signicantincrease in enrollment rom 2000 to 2008wereArizona,Florida,Georgia,Nevada,NorthCarolina, Texas, and Utah. The states whoseincrease was not statistically signicant wereColorado,Delaware,Idaho,SouthCarolina,Tennessee,andVirginia.

from60to72percentfrom2000

to 2008.

In2008,41percentofstudents

enrolled in ull-day kindergarten

programs had a mother employed

ull-time, compared with the15percentwhosemothers

were employed part-time, and

32 percent o ull-day kindergarten

students whose mothers were not

in the labor orce.9Oneinfourstu-

dents enrolled in ull-day programs

came rom amilies with incomes o

$75,000 or more.

PRIMARY AND SECONDARY

ENROLLMENT

The ACS is used in this sectionbecause it provides a look at

geographic variations in school

enrollment.Overallenrollment

inGrades1through12fellslightly

9Theremaining13percentofchildrenin ull-day kindergarten programs were madeup o children not living with a mother(7 percent) and children whose mothers wereunemployed(6percent)andthesewerenotstatistically dierent rom one another.

enrolled in ull-day programs,

comparedto26percentof

non-Hispanic White kindergart-

ners.In2008,81percentofBlack

students were enrolled in ull-day

kindergarten programs, compared

to 70 percent o non-Hispanic

Whitestudentsand69percentof

Hispanic students.

The increase in attendance o ull-

day kindergarten programs may

refect the changing needs o chil-

dren and their amilies. However,

changes in kindergarten enrollment

patterns are not perectly associ-

ated with changes in employment

patternsoftheirparents.In1986,

57 percent o married coupleswithchildrenunder18hadboth

spouses in the labor orce. The

ercentageroseto68percentin

2000 and was slightly lower at

66percentin2008.Thepercent-

age o children enrolled in ull-day

kindergarten continued to increase

0

10

20

30

40

50

60

70

80

90

100

2008200319981993198819831978

Figure 3.

Enrollment in Full- and Part-DayKindergarten: 1978 to 2008

Source: U.S. Census Bureau, Current Population Survey, October 19782008.

For information on sampling and nonsampling error, see

.

Proportion part-day

Proportion full-day


6/20


grade variable was calculated using

enrollment levels by age and the

month the survey was asked.13

Enrollmentbelowmodalgradein

school could be due to late entry

into school or to repeating grades

ater entering school. Most grade

retention occurs at kindergarten,

rst and ninth grades.14 Some

amilies choose to wait until their

childrenare6yearsoldtostart

13 More specically, modal grade wascreated by assigning a grade level based onmonth o interview and birth year and month.Calculations were based on two groups: thoseinterviewed between January and September,andthoseinterviewedbetweenOctoberandDecember.

14Datafromthe2004OctoberCPSshowed that grade retention was most com-monatkindergarten(12percent),rstgrade(20percent),andninthgrade(13percent).There was no statistical dierence betweenkindergarten and ninth grade.

according to their relative prog-

ress in school: that is, whether the

grade or year in which they were

enrolled was below, at, or above

the typical (modal) grade or people

o their age at the time o the sur-

vey. The modal grade is the year o

school in which the largest propor-

tion o students o a given age is

enrolled. For example, modal grade

for15-year-oldsistenthgrade.

Because o the sampling strategy o

the ACS, it was necessary to create

a modal grade variable with slightly

dierent properties rom that tra-

ditionally reported using CPS data.

The ACS asks respondents i they

were enrolled at any time in the

past 3 months and this is asked or

a dierent month o the year or

dierent respondents. The modal

17-year-olds(28percent).Forthe

majorityofstates,however,

overall changes in population and

enrollment went together. States

with high population growth and

high enrollment growth included

Arizona,Utah,andGeorgia.12

States with decreases in both

population and enrollment included

NorthDakota,Vermont,Louisiana,

and Maine.

MODAL GRADE

The ACS is used in this section

because o its geographic scope

and size. For the rst time in 2008,

the ACS provides inormation on

single grades o enrollment. This

report classies enrolled people

12 Nevada also had high population andenrollment growth.

20.0 or over

10.0 to 19.9

0.1 to 9.9

10.0 to 0.0

Under 10.0

MT

AK

NM

OR MN

KS

SD

ND

MO

WA

FL

IL IN

WI NY

PA

MI

OH

IA

ME

MA

CT

AZ

NV

TX

COCA

WY

UT

ID

NE

OK

GA

AR

AL

NC

MS

LA

TN

KY

VA

SC

WV

RI

DE

MD

NJ

HI

VTNH

DC*

Percentage change

Figure 4.

Percentage Change in Enrollment, Grades 1 through 12:2000 to 2008

Source: U.S. Census Bureau, Census 2000 and 2008 American Community Survey. For information on sampling and nonsampling

error, see .


7/20


Lookingatamapofenrollment

belowmodalgradefor6-to

9-year-olds shows the contrastbetween the central states and

statesalongtheEastandWest

Coasts (Figure 5). Coastal states

had lower percentages o enroll-

ment below modal grade than the

states in the middle o the country.

The maps o enrollment below

modalgradefor10-to13-year-olds

and14-to17-year-oldsbystate

show how the proportion o chil-

dren below modal grade increases

withage(Figures6and7).Enroll-

ment below modal grade was high-

est in the South, upper plains, and

Midwestern states. The net result

is that over hal o the states have

between15and30percentoftheir

14-to17-year-oldstudentsenrolled

below modal grade.

grade than emale students (22

percentcomparedwith16percent)

and accordingly, there were moreemales enrolled at modal grade

thanmales(67comparedwith

53 percent).

Black students had the highest

percentageof6-to17-year-olds

enrolled below modal grade

and Asian students had the lowest

percentage o enrollment below

modal grade (22 percent and

12percent,respectively).Foreign-

born students were more likely

than native-born students tobe enrolled below modal grade

(23 percent compared with

19percent)andenrolledabove

modalgrade(21percentcompared

with16percent).Disabledstudents

were more likely than students

without a disability to be enrolled

below modal grade (34 percent

comparedwith18percent).

kindergarten to ensure that their

children are mature enough, either

physically or academically, orschool. Sometimes this phenom-

enonisreferredtoasacademic

redshirting.15

In2008,themajorityofstudents

aged6to17(65percent)were

enrolled at their modal grade, with

19percentofstudentsenrolled

belowmodalgrade,and16per-

cent enrolled above modal grade

(Table 2). Male students were more

likely to be enrolled below modal

15GraueandDiPernausedthetermacademicredshirtingtorefertochildrenwho delay entrance into kindergarten bya year, analogous to the college ootballpractice o redshirting, which means aplayer has participated in the academic yearbut not that years sports season, extendingtheir eligibility to play sports or anotheryear.Graue,M.ElizabethandJamesDiPerna,RedshirtingandEarlyRetention:WhoGetstheGiftofTimeandWhatAreItsOutcomes?American Educational Research Journal,No. 37, 2000, pp. 509534.

Table 2.

Modal Grade Status by Selected Characteristic for Students Aged 6 to 17: 2008(Populationaged6to17years.Numbersinthousandsandpercents)

Characteristic

Total

Modal grade status1

Below modal grade At modal grade Above modal grade

Total 47,793 189 647 164SexMale 24,443 217 528 155Female 23,349 160 666 175

Race and Hispanic OriginWhite alone 33,431 188 663 150White alone, non-Hispanic 27,532 187 675 138

Black alone 6,957 219 587 194Asian alone 1,891 120 652 227

Hispanic (any race) 9,612 187 606 207

NativityNative-born 45,503 187 651 162Foreign-born 2,289 232 558 210

DisabilityNot disabled 45,255 181 652 167Disabled 2,537 337 543 120

Language Spoken at HomeEnglish only 38,095 192 657 150Language other than English 9,697 177 604 219

1 The modal grade is the year o school in which the largest proportion o students o a given age is enrolled For more inormation, read the section under

Modal Grade

Source: US Census Bureau, 2008 American Community Survey For inormation on sampling and nonsampling error, see


8/20


Resultsfromourbasicmodelreveal

that compared to girls, boys had

a higher likelihood o enrollment

belowmodalgrade(Model1,Table

3).Enrollmentbelowmodalgrade

also increased with age. Children

with a disability had a much higher

probability o enrollment below

modal grade than children without

a disability. Similarly, children living

in poverty had a higher probabil-

ity o being enrolled below modalgrade than children not living

in poverty.

The regression included some

interaction terms. These show how

some eects act in combination.

Forexample,theoddsof1.52for

Maleinthemodelshowthatthe

probability o being below modal

control o school, language spoken

at home, and region o residency.

Table 3 displays results rom logis-

tic regression analysis predicting

enrollment below modal grade by

selectedcharacteristics.Results

are displayed as odds ratios, which

are related to the probability o

enrollment below modal grade,

ater allowing or the infuence o

othervariablesinthemodel.Values

above1indicatethat,comparedto

the reerence group, children have

a higher probability o enrollment

below modal grade, net o other

variablesinthemodel.Valuesless

than1indicatethat,comparedto

the reerence group, children have

a lower probability o enrollment

below modal grade.

MULTIVARIATE ANALYSIS

A multivariate analysis was per-

ormed to ascertain the indepen-

dent eects o certain character-

istics on the likelihood o being

enrolled below modal grade. The

model allows us to examine the

eects o characteristics such as

sex, race, and disability while

controlling or other actors (or

example, age, nativity, and region)

that infuence the probability o

being below modal grade. Fac-

tors in the model include sex, age,

nativity, disability, poverty,16

16 A child was counted as living in povertyi the amily in which he lived had incomelower than the ocial poverty threshold. Formore inormation, see .

25.0 and over

15.0 to 24.9

Under 15.0

U.S. average: 11.2 percent

MT

AK

NM

ORMN

KS

SD

ND

MO

WA

FL

IL IN

WI NY

PA

MI

OH

IA

ME

MA

CT

AZ

NV

TX

COCA

WY

UT

ID

NE

OK

GA

AR

AL

NC

MS

LA

TN

KY

VA

SC

WV

RI

DE

MD

NJ

HI

VTNH

DC*

Percent

Figure 5.

Percentage Below Modal Grade, Aged 6 to 9: 2008

Source: U.S. Census Bureau, 2008 American Community Survey. For information on sampling and nonsampling error,

see .


9/20


students are more likely to be

below modal grade. The main

reason or this dierence is that the

eect o being below modal grade

varied across regions o the coun-

try. Students in the Midwest and

in the South were more likely to

be below modal grade than those

intheWest.Inourrstregression,

the eect o Hispanic origin was

obscured by the act that many

Hispanics are located in the West,

where ewer students are belowmodal grade overall.

Oursecondmodelwasonein

which we examined how the eect

o race and Hispanic origin varied

byregion(Model2,Table3).Inthe

model, the odds o being below

modal grade or Black (.84) and

have a larger eect at later ages.

Boys with a disability have a much

higher probability o enrollment

below modal grade compared with

boys without a disability. The eect

o poverty on enrollment below

modal grade is much higher at

age16thanatage6.

Amajoradvantageofthe

regression model or enrollment

below modal grade is that we

can closely examine the eectsofraceandHispanicorigin.Inour

tabular analysis presented earlier

(Table 2), Hispanic students were

ound to be no more likely than

non-Hispanic White students to

be below modal grade. By contrast,

our rst regression model (Model

1,Table3)showsthatHispanic

grade is higher or boys than or

girls in the reerence age category.

ThevalueofoddsforMale*Age

isalsogreaterthan1,indicating

that the probability o being below

modal grade increases aster with

age or boys than or girls. This

can be seen by examining the

predicted probabilities17 or

enrollment below modal grade

(Figure 8). The eects o disability

and poverty on enrollment below

modal grade are cumulative and

17Predicted probabilities are the probabili-ties expected or given types o people, basedon a regression model, such as the regressiono enrollment below modal grade shown inTable 3. The predicted probability or emalesage6,forexample,canbecomparedwiththepredictedprobabilityforfemalesage16withdierences attributable to poverty, nativ-ity, language, type o school, race, Hispanicorigin, and region all held constant.

25.0 and over

15.0 to 24.9

Under 15.0

MT

AK

NM

ORMN

KS

SD

ND

MO

WA

FL

IL IN

WI NY

PA

MI

OH

IA

ME

MA

CT

AZ

NV

TX

COCA

WY

UT

ID

NE

OK

GA

AR

AL

NC

MS

LA

TN

KY

VA

SC

WV

RI

DE

MD

NJ

HI

VTNH

DC*

Percent

Figure 6.



see .


10/20


predicted probabilities, which show

the eect o combining these terms

together (Figure 9).18

All three race and Hispanic origin

groups examined in Figure 9 show

lower probability o being below

modal grade in the Northeast and

the West than in either the South

or the Midwest (or each race and

Hispanic origin group, all our

estimates or regions are signi-

cantly dierent rom one another).As with Hispanic students, the

probability o Black students being

below modal grade in the West is

smaller than that o non-Hispanic

18The predicted probabilities in Figure9areformales,age16,notinpoverty,notdisabled,inpublicschool.Race,Hispanicorigin, language use, and oreign-born statusare listed in the gure.

White students. Black students in

the Midwest are also less likely

than non-Hispanic White students

to be below modal grade. By con-

trast, non-Hispanic White students

are less likely to be enrolled below

modal grade in the Northeast and

in the South. Hispanic students

are not signicantly dierent rom

non-Hispanic White students in

the Midwest.

Note that in making thesecomparisons, we have controlled

or important characteristics that

aect the probability o being

enrolledbelowmodalgrade.In

particular, we are examining eects

o Hispanic origin, net o immigra-

tion and language actors. Hispanic

students born outside the United

Hispanic (.93) students represent

what is happening in the West,

since that is the reerence category

or interactions. That these odds

aresignicantlybelow1indicates

that in the West, Black and Hispanic

students are less likely to be below

modal grade than the reerence

non-Hispanic White group. For

other regions o the country, the

odds o being below modal grade

relative to non-Hispanic White

students is measured by combin-

ing the basic odds o being below

modal grade (or example, .84 or

Black students) and the interaction

o race and Hispanic origin with

region(forexample,1.41forBlack

studentsintheNortheast).Inorder

to interpret these interactions, it

is easiest to turn once again to

25.0 and over

15.0 to 24.9

Under 15.0

MT

AK

NM

OR MN

KS

SD

ND

MO

WA

FL

IL IN

WI NY

PA

MI

OH

IA

ME

MA

CT

AZ

NV

TX

COCA

WY

UT

ID

NE

OK

GA

AR

AL

NC

MS

LA

TN

KY

VA

SC

WV

RI

DE

MD

NJ

HI

VTNH

DC*

Percent

Figure 7.



see .


11/20


States and who speak a language

otherthanEnglishathomeare

more likely than non-Hispanic White

students to be below modal grade

in all regions o the country

(Figure 9).

DROP-OUTS

The Census Bureau calculates two

types o drop-out rates. The event

drop-out rate was rst introduced

intheOctober1983CPSSchool

Enrollmentreportasawayto

approximate longitudinal data

by asking persons what their

enrollment status currently is,

and what it was at some dened

point in the past. The CPS asks a

retrospective question on enroll-

ment,Wereyouenrolledinschool

lastOctober?(1yearago).Based

on the answers to this question

and the current enrollment item,

the numbers o persons who were

enrolled a year ago, but who pres-

ently are not, are estimated. The

event drop-out rate captures the

percentageofyouthaged15to

24whodroppedoutofGrades

10through12inthe12months

betweenoneOctoberandthenext

(October2007andOctober2008).

In2008,3.3percentof15-to

24-year-olds who had been enrolled

inGrades10through12theprevi-

ous year had dropped out o school

(event drop-outs, Table 4).19 The

high school event drop-out rate was

lowerforstudentsaged15to17

(3 percent) than or students aged

19Asianstudents(4percent)and18-to

19-year-old(4percent)drop-outrateswerenot statistically dierent rom the total o15-to24-year-olddrop-outs(3percent).

Table 3.

Regression Models, Enrollment Below Modal Grade: 2008

Characteristic

Odds ratio o enrollmentbelow modal grade

Model 1 Model 2

SexFemale (R) (R)

Male *152 *152

NativityNative-born (R) (R)Foreign-born *142 *142

DisabilityNot disabled (R) (R)Disabled *207 *207

Language Spoken at HomeEnglish only (R) (R)Language other than English *087 *087

PovertyNot in poverty (R) (R)In poverty *179 180*

Type of SchoolPublic (R) (R)Private 100 *101

RegionWest (R) (R)Midwest *160 *154South *159 *143Northeast 103 *092

Race and Hispanic OriginWhite alone, non-Hispanic (R) (R)Black *104 *084

Hispanic (any race) *108 *093

Other race *083 *082

Race and Hispanic Origin InteractionsWith Region

Black alone:West (X) (R)Midwest (X) 104South (X) *132Northeast (X) *141

Hispanic (any race): (X)West (X) (R)Midwest (X) 103South (X) *130Northeast (X) *135

(R) Reerence category

* Signifcant at 001 level

(X) Not applicable

Source: US Census Bureau, 2008 American Community Survey For inormation on sampling andnonsampling error, see


12/20


20 to 24 (9 percent).20 Females had

a higher high school drop-out rate

than males, 4 percent compared

with 3 percent. Non-Hispanic White

students (2 percent) had a lower

drop-out rate than Black students

(6percent)andHispanicstudents

(5 percent).21

The second measure o drop-outs

used by the Census Bureau is

thedrop-outpool,whichisthe

percentageofyouthaged15to

24 who are not enrolled in school

and who do not have a high school

degree (regardless o when they

might have let school).

In2008,7percentof15-

to 24-year-olds were in the

drop-out pool. The number varies

by age because younger studentsare less likely to be high school

graduates.Threepercentof15-

to17-year-olds,9percentof18-

to19-year-olds,and10percent

o 20- to 24-year-olds were not

enrolled and were not high school

graduates. Hispanic students

were more likely to be in the

drop-out pool than other students,

17percentcomparedto5percent

o non-Hispanic White students,

4 percent o Asian students, and9 percent o Black students.22

COLLEGE ENROLLMENT

The CPS is used in this section to

discuss college enrollment because

it provides data on type o school

(2- or 4-year) and type o atten-

dance(full-orpart-time).In2008,

20 The high school event drop-out rate wasnot statistically dierent between age groupsof15-to17-year-oldsand18-to19-year-olds.

21 The event drop-out rate o Whitestudents (2 percent) was not statisticallydierent rom Asian students (4 percent).The event drop-out rate o Black students(6percent)wasnotstatisticallydierentrom Hispanic students (5 percent) or Asianstudents (4 percent). The event drop-outrate o Asian students (4 percent) was notstatistically dierent rom Hispanic students(5 percent).

22 The percent o non-Hispanic Whitestudents in the drop-out pool was not sta-tistically dierent rom the percent o Asianstudents in the drop-out pool.

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

Foreign-born,Hispanic, speaks

language other thanEnglish at home

Hispanic(of any race)

BlackNon-HispanicWhite

Figure 9.

Predicted Probabilities of Enrollment Below ModalGrade by Region, Race, Hispanic Origin,and Nativity: 2008

Note: The predicted probablilities shown here are based on average effects for peoplewho vary by the characteristics shown (race, Hispanic origin, nativity and speaking alanguage other than English at home, age, sex, poverty, and disability). It is assumedthat the people being compared are male, age 16, not in poverty, and attend publicschools.

Source: U.S. Census Bureau, 2008 American Community Survey. For information onsampling and nonsampling error, see .

South

Probability

Total MidwestNortheast West

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

Male, belowpoverty

Male, withdisability

MaleFemale

Figure 8.

Predicted Probabilities of Enrollment Below ModalGrade by Age, Sex, Disability, and Poverty: 2008

Note: The predicted probabilities shown here are based on average effects for peoplewho vary by the characteristics shown (age, sex, poverty, and disability). It is assumed

that the people being compared are native-born, non-Hispanic White, from the sameregion (West), all speak only English at home, and attend public schools.

Source: U.S. Census Bureau, 2008 American Community Survey. For information onsampling and nonsampling error, see .

Age 6

Age 16

Probability


13/20


Table 5.

Enrollment in Undergraduate and Graduate College by Selected Characteristic: 2008(Numbers in thousands and percents)

Characteristic

College enrollment

Total

2-year institution 4-year institution

Total Full-time Part-time Total Full-time Part-time

Total 18,596 5,793 192 120 12,802 519 170

SexMale 8,287 2,488 189 112 5,799 545 155Female 10,309 3,305 194 127 7,003 498 181

Age16 to 18 years 2,156 884 350 60 1,271 575 1519 to 20 years 4,239 1,488 284 68 2,751 621 2821 to 22 years 3,220 659 141 64 2,562 728 6823 to 24 years 2,055 489 146 92 1,566 594 16825 years and older 6,925 2,272 123 205 4,652 320 352

Race and Hispanic OriginWhite alone 14,376 4,427 193 115 9,950 519 174

White alone, non-Hispanic 12,300 3,490 181 103 8,809 543 174Black alone 2,478 895 212 149 1,583 469 170Asian alone 1,217 311 136 119 906 615 129

Hispanic (any race) 2,222 998 262 187 1,224 382 169

EmploymentFull-time 5,533 1,861 102 233 3,693 260 405Part-time 5,240 1,753 246 89 3,487 575 91Not employed 7,802 2,180 219 60 5,622 666 55

Source: US Census Bureau, Current Population Survey, October 2008 For inormation on sampling and nonsampling error, see

Table 4.

Event Drop-Out Rates and High School Noncompleters, Aged 15 to 24: 2008(Numbers in thousands and percents)

Characteristic

Population 15 to 24 years old

Total

Enrolled in grades 1012in the previous year

No high school completion andnot enrolled in school

Total Dropped out1 Eventdrop-out rate Number Percent

Total 41,696 11,750 390 33 3,047 73

SexMale 21,062 5,999 174 29 1,624 77Female 20,635 5,751 216 38 1,423 69

Age15 to 17 years 12,746 7,051 192 27 345 2718 to 19 years 8,492 4,146 146 35 750 8820 to 24 years 20,459 553 51 93 1,951 95

Race and Hispanic OriginWhite alone 32,188 8,942 246 28 2,241 70White alone, non-Hispanic 25,374 7,079 156 22 1,129 45

Black alone 6,259 1,868 114 61 576 92Asian alone 1,620 429 17 39 64 40

Hispanic (any race) 7,498 2,062 101 49 1,236 165

1 Dropped out means not enrolled and not a high school graduate, given that the person was enrolled in high school the previous year

Source: US Census Bureau, Current Population Survey, October 2008 For inormation on sampling and nonsampling error, see


14/20


SchoolEnrollmentSupplementto

theOctober2008CPSisthecivil-

ian noninstitutionalized population

living in the United States. The

institutionalized population, which

is excluded rom the population

universe, is composed primarily

o the population in correctional

institutions and nursing homes

(91percentofthe4.1million

institutionalized people in

Census 2000).

ACCURACY OF THEESTIMATES

Statistics rom sample surveys

aresubjecttosamplingerrorand

nonsampling error. All comparisons

presented in this report have taken

sampling error into account and

are signicant at the 90 percent

condence level. This means the

90 percent condence interval or

the dierence between estimates

being compared does not include

zero. Nonsampling error in surveys

may be attributed to a variety o

sources, such as how the survey

was designed, how respondents

interpret questions, how able and

willing respondents are to provide

correct answers, and how accu-

rately answers are coded and clas-sied. To minimize these errors,

the Census Bureau employs qual-

ity control procedures in sample

selection, the wording o questions,

interviewing, coding, data process-

ing, and data analysis.

The nal ACS population estimates

areadjustedintheweightingpro-

cedure or coverage error by con-

trolling specic survey estimates to

independent population controls by

sex, age, race, and Hispanic origin.

This weighting partially corrects or

bias due to over- or undercoverage,

Inallagegroups,exceptthose

aged25andover,themajorityof

students were enrolled ull-time at

4-yearcolleges.Ofstudentsaged

25 and over, more than hal were

enrolled part-time at both 2-year

and4-yearinstitutions(56percent).

In2008,overhalfofcollegestu-dents (55 percent) were women.

Womenmadeup56percentof

enrollment at 2-year colleges and

55 percent o enrollment at 4-year

colleges. Women also made up the

majorityofgraduatestudents

(57 percent).24

SOURCES OF THE DATA

Most estimates in this report

are rom the 2008 American

Community Survey (ACS), the2008Octobersupplementtothe

Current Population Survey (CPS),

and Census 2000. Some estimates

are based on data obtained by the

CPS in earlier years.

The population represented

(the population universe) in

theOctober2008ACSincludes

both the household and the group

quarters populations (that is, the

resident population). The group

quarters population consists o

the institutionalized population

(such as people in correctional

institutions or nursing homes)

and the noninstitutionalized

population (most o who reside

in college dormitories).

The population represented

(the population universe) in the

24 The percentage o students enrolledthat are emale is not statistically dier-

ent rom its subgroup o the percentage ostudents enrolled in 2-year colleges that areemale. Additionally, the percentage o emalestudents enrolled in 2-year colleges was notstatistically dierent rom the percentage oemale students enrolled in 4-year colleges orgraduate school.

college enrollment was at an all-

timehighwith18.6millionstu-

dents enrolled, up 22 percent rom

15.3millionin2000(Table5).

Contributing to the overall increase

in college enrollment in 2008 was

an all-time high in 2-year college

enrollment, with 5.8 million stu-dents enrolled, up 38 percent rom

3.9 million in 2000. As a conse-

quence, because the enrollment in

4-year colleges did not increase at

that rate, the percentage o college

students enrolled in 4-year colleges

has declined since 2000, with more

students attending 2-year col-

leges. Sixty-nine percent o college

students were enrolled at a 4-year

college in 2008, down rom

72 percent in 2000.Full-time enrollment is on the

rise or students attending both

4-and2-yearcolleges.In2008,

75 percent o students at 4-year

colleges attended ull-time, up

from71percentin2000.At2-year

collegesin2008,61percentof

students were enrolled ull-time,

up rom 55 percent in 2000.

Contributing to the increase o

enrollment at 2-year colleges was

an increase in Hispanic college

enrollment.In2008,1million

Hispanic students were enrolled in

2-year colleges, up 85 percent rom

540,000 in 2000. A higher num-

ber o Hispanic college students

were enrolled in 2-year colleges

(45 percent) compared with other

groups:Blackstudents(36per-

cent), non-Hispanic White students

(28 percent), and Asian students

(26percent).23

23 Non-Hispanic White students(28 percent) were not statistically dierentfromAsianstudents(26percent).


15/20


intervals, can be ound at

or by contact-

ing Dave Hall o the Demographic

Statistical Methods Division via

e-mail at .

MORE INFORMATION

Detailed tabulations, related inor-

mation, and historic data are avail-

ableontheInternetattheSchool

EnrollmentpageontheCensus

Bureaus Web site at .

For additional questions or com-

ments, contact Jessica W. Davis

at301-763-2464orviae-mailat

.

but biases may still be present, or

example, when people who were

missed dier rom those inter-

viewed in ways other than sex, age,

race, and Hispanic origin. How this

weighting procedure aects other

variables in the survey is not pre-

cisely known. All o these consid-

erations aect comparisons across

dierent surveys or data sources.

For inormation on sampling and

estimation methods, condential-

ity protection, and sampling and

nonsampling errors, please see the

2008ACSAccuracyoftheData

document located at

.

The CPS weighting procedure uses

ratio estimation whereby sample

estimatesareadjustedtoinde-

pendent estimates o the national

population by age, race, sex, and

Hispanic origin. This weighting

partially corrects or bias due to

undercoverage, but biases may still

be present when people who are

missed by the survey dier rom

those interviewed in ways other

than age, race, sex, and Hispanic

origin. How this weighting proce-

dure aects other variables in the

survey is not precisely known. All

o these considerations aect com-

parisons across dierent surveys

or data sources. Further inorma-

tion on the source o the data and

accuracy o the estimates, includ-

ing standard errors and condence


16/20


Appendix Table A.

Comparison of Census Bureau Data Sources on School Enrollment

Survey Characteristics American Community Survey Current Population Survey

Geographic scope Annual est imates o the nation, regions, states,congressional districts, and geographies o 65,000or more Three-year estimates available or places o20,000 or more and 5-year estimates o areas as small

as census tracts

National estimates and estimates o selectedcharacteristics or regions and states

Periodicity o collection Every year Every year

Timeliness Released year ater collection cycle Released ater year o collection cycle

Sample size Annual sample o about 3 million addresses Data arecollected rom about one-twelth o the sample eachmonth

The school enrollment data come rom October CPSsupplement, which is based on a sample o about72,000 addresses

Data collection method Mail, telephone, and personal-visit interviews or the 50states, the District o Columbia, and Puerto Rico Abouthal the responses are obtained by mail The ACS is amandatory survey

Telephone and personal-visit interviews or the 50 statesand the District o Columbia The CPS is a voluntarysurvey

Questionnaire item(s) The ACS asks respondents every month, At any timein the last 3 months, has this person attended regularschool or college? The respondent is asked what grade

or level this person attended

In the month o October, the CPS asks a detailed andextensive list o questions about school enrollment in thecurrent and past year There are separate questions or

adults 15 years and older and or children 314 yearsold

Unique measures/data ACS enrollment data can be produced at the nationallevel and very small levels o geography

The CPS provides data regarding single-year enroll-ment, enrollment status and level or the previous year,whether the respondent goes to school on a ull-time orpart-time basis, whether they attend a 2-year or a 4-yearinstitution, whether they are obtaining any vocationaltraining, what year they received their postsecondarydegree, and whether they received a GED (generaleducational development test-based certifcation)

Technical issues ACS statistics on school enrollment are based oninterviews conducted during the entire year

CPS statistics on enrollment are based on interviewsconducted in October

Population universe The ACS includes the resident population o theUnited States, including household and group quarterspopulations

The CPS includes the civilian noninstitutionalizedpopulation and Armed Forces personnel living o postor with their amilies on post

Tables available/detail ACS school enrollment tables can be accessed throughAmerican FactFinder showing school enrollment or thenation and smaller geographies by characteristics suchas age, sex, educational attainment, and poverty status

Detailed table package consists o eight tables withschool enrollment or the nation shown by character-istics such as age, race, sex, amily income, type ocollege, employment status, and vocational courseenrollment

Sampling error inormation Only or published tables Can be computed by data user

Historical data The ACS began in 1996 in a limited number o test sitesand began national implementation in 2000

Enrollment data rom the CPS have been gatheredsince 1947

Public-use fle Yes Yes

Electronic accessibility TablesAmerican FactFinderPublic-use flesDataFerret

TablesSchool Enrollment home pagePublic-use flesDataFerret


17/20


18/20


19/20


20/20

U.S. Department o CommerceEconomicsandStatisticsAdministration

U.S.CENSUSBUREAU

Washington, DC 20233

OFFICIAL BUSINESS

Penalty or Private Use $300

FIRST-CLASSMAILPOSTAGE&FEESPAIDU.S. Census Bureau

PermitNo.G-58

Date post:	07-Apr-2018
Category:	Documents
Upload:	source-media-group-news
View:	216 times
Download:	0 times

Census Report on School Enrollment

Documents