2012 Applied Demography Conference Session 2C Julien Bérard-Chagnon Demography Division Statistics Canada Monday, January 9, 2012 Using tax data to estimate.

2012 Applied Demography ConferenceSession 2C

Julien Bérard-ChagnonDemography Division

Statistics Canada

Monday, January 9, 2012

Using tax data to estimate the number of families and

households in Canada

Statistics Canada • Statistique Canada2

Context

Households (HHs) estimates are used in the calibration of many of Statistics Canada’s surveys

Census families (CFs) estimates are released yearly

Limitations in previous methodologies fuelled the need to develop a new, integrated methodology

3

Plan of the presentation

Concepts Previous methodologies T1 Family File (T1FF) New methodology Comparisons Concluding remarks

Statistics Canada • Statistique Canada

Statistics Canada • Statistique Canada4

Concepts

Census families• Married or common-law couple • Lone-parent family

Households• Person or a group of persons who occupy the same

dwelling

Statistics Canada • Statistique Canada5

Previous methodologies

Census families• Number : Component-cohort approach• Characteristics : extrapolation based on previous

censuses

Households• Headship ratios based on previous censuses

Significant limitations

6

T1 Family File (T1FF)

T1 income tax data from the Canada Revenue Agency (CRA)

Aims at recreating Canadian population and families with tax data

Produced yearly since 1982 Very good coverage Contains basic demographic information

Statistics Canada • Statistique Canada

7

New method

T1FF heads

1. Coverage adjustment

2. Bias adjustment

3. Period adjustment

4. Census coverage adjustment

Statistics Canada • Statistique Canada

)(),,(

),,()()()(

.)(

.)(1 XCNU

sapP

sapPXYCHXEST tiCNUDem

CDnt

CNUDemJulynt

tFFTntCFnti

(1) (2) (3) (4)

8

1. Coverage adjustment

Adjust for coverage with population estimates

Done every year ),,(

),,()()( 1

.11

sapP

sapPYHYCH FFT

t

DemtFFT

tFFT

t

Statistics Canada • Statistique Canada

15 to 24 years

25 to 34 years

35 to 44 years

45 to 54 years

55 to 64 years

65 to 74 years

75 years and over

90%

94%

98%

Coverage by age group, Canada, 2006

)()( 1 YCHXEST FFTntCFnti

9

2. Bias adjustment

Comparison of T1FF and last Census

Assumed to be constant until next Census From CFs to HHs : Census CF-HH relationship

)(

)()

1

.

YCH

XHX

FFT

Cens

Statistics Canada • Statistique Canada

)()()( 1 XYCHXEST tFFTntCFnti

10

3. Period adjustment

Desired date of reference : July 1st

date of reference : Census Day (CD) Headship ratios

• Exclusion of Census net undercoverage (CNU) from population estimates

)()(1 XYCH FFTnt

Statistics Canada • Statistique Canada

),,(

),,()()()(

.)(

.)(1

sapP

sapPXYCHXEST

CNUDemCDnt

CNUDemJulynt

tFFTntCFnti

11

4. Census coverage adjustment

Method uses Census coverage studies to estimate CF net undercoverage• HH net undercoverage is obtained with Census ratios

Assumed to be constant until next Census• 2006 : around 150,000 CFs and 335,000 HHs

Statistics Canada • Statistique Canada

)(),,(

),,()()()(

.)(

.)(1 XCNU

sapP

sapPXYCHXEST tiCNUDem

CDnt

CNUDemJulynt

tFFTntCFnti

Comparison by size (with the 2006 Census)

Households

Size 1

Size 2

Size 3+

Total

-8% -6% -4% -2% 0% 2%

Previous New

Census families

Size 2

Size 3

Size 4+

Total

-8%-6%-4%-2% 0% 2% 4% 6%

Previous New

12 Statistics Canada • Statistique Canada

Dynamism of the estimates

Statistics Canada • Statistique Canada13

HHs (left) and Unemployment rate (right),

15 to 24 years

2006 2007 2008 2009 2010 2011495,000500,000505,000510,000515,000520,000525,000530,000535,000540,000545,000

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

15 to 24 years Unempl. 15 to 24 years

14

Concluding remarks

New method based on tax data Internal consistency between CFs and HHs Dynamic estimates Better results

Statistics Canada • Statistique Canada

15

Questions or comments ?

You can consult our population and family estimation methods:• Internet : www.statcan.gc.ca

Contact us :• Client Services:

1-866-767-5611• Email :

[email protected] Thank you very much ! Merci beaucoup !

Statistics Canada • Statistique Canada