Introduction

An income quintile is a measure of neighbourhood socioeconomic status that divides the population into 5 income groups (from lowest income to highest income) so that approximately 20% of the population is in each group. At MCHP, we have created income quintiles for two distinct population groups: urban (Winnipeg and Brandon) and rural (other Manitoba areas).

This concept contains information in the following sections:


• data sources required for generating income quintiles;
  
• the steps for creating Income Quintile groups;
  
• postal codes that cannot be ranked and the Income Not Found group;
  
• a comparison of methodologies for the Standard Income Quintile versus the Child Health Income Quintile;
  
• urban and rural income quintile considerations;
  
• income quintile values and ranges;
  
• SAS® code and formats; and
  
• a variation on the Child Health Income Quintile Methodology.

Data Sources

In order to generate Income Quintiles for socioeconomic analyses, the following data sources are required. The purpose of each data source is explained:


1. the Manitoba Health Insurance Registry Data is used to generate a population file for a selected year.
   
2. the public-use Canada Census data files developed by Statistics Canada are used to identify the average household income value by Dissemination Area (DA).
   
3. the Postal Code Conversion File (PCCF) developed by Statistics Canada is used to identify and verify postal codes that fit within Dissemination Areas (DA).
   
4. the Manitoba Health Population File is used to identify the population of Manitoba at the time the income quintiles are generated.

Steps for Creating Income Quintiles Groups

Income Quintile information can be created for each year using the following general steps:


1. generate the population file for a selected year.
   
2. remove the postal codes that cannot be ranked. (see the next section for more information on why postal codes cannot be ranked).
   
3. attach the average household income value from the Census files to the population file using the Postal Code Conversion File (PCCF).
   
4. rank the population by Urban/Rural geographical location and by average household income.
   
5. form the 20% population income quintile groups based on the average household income values and population count.

The data will be coded into one of the following groups:


• Urban Income Quintiles: U1 (lowest) to U5 (highest);
  
• Rural Income Quintiles: R1 (lowest) to R5 (highest); or
  
• Income Not Found (NF).

If the postal code cannot be ranked to an income quintile the value will be Not Found (NF).

Postal Codes That Cannot Be Ranked

There are several reasons why a postal code cannot be ranked. These include:


• greater than 90% of the population of the same postal code are associated with a Personal Care Home (PCH).
  
• the postal code is associated with a mental health institution, a prison, the Public Trustee Office, or a Child and Family Services (CFS) office.
  
• the postal code is not on the PCCF.
  
• the postal code is on the PCCF but the average household income value in the Census data file is suppressed.
  
• the postal code does not appear on the population file.

Data in the population file that contains postal codes that cannot be ranked are categorized as Income Unknown (Income Not Found (NF)) and are excluded from the Income Quintile analyses.

Example: Breakdown of the Income Not Found Group for the 2006 Population

The following table provides a breakdown (percentage) of the "reasons for unranked postal codes" and the total percent of the population that is in the Income Not Found grouping for the 2006 population. This analysis uses the 2008 Postal Code Conversion File (PCCF) to match to postal codes in the population file. This information represents a specific "snapshot" in time and would be different if a different PCCF was used.

Reason for Unranked Postal Code	2006
Postal code not on PCCF/Population File	11.52%
Unrankable Postal Code (prisons, Public Trustee, Family Services)	42.43%
PCH Resident	40.48%
Suppressed Income	5.57%
Total of Income Not Found	100%
Percent of Total Population with Income Not Found	0.84%

NOTE: In our analyses, the percent of the population within each of the Income Not Found categories will vary depending on the PCCF and Population files that are used, but overall, the Income Not Found group consistently represents a very small portion of the entire population.

Comparing Methodologies: The Standard Income Quintile (SIQ) versus The Child Health Income Quintile (CHIQ)

MCHP first started using Income Quintiles as a measure of socio-economic status (SES) in the mid 1990's. The methods used to calculate the Standard Income Quintile (SIQ) are described in the concept Income Quintiles Based on the 1996 Census. The method used to calculate Income Quintiles was updated in 2001 due to changes in the 2001 Census and was first implemented for the Assessing the Health of Children in Manitoba: A Population Based Study deliverable. This new method is referred to as the Child Health Income Quintile (CHIQ). The SIQ method differs from the CHIQ method in two significant ways:


1. The SIQ values are based on enumeration area (EA) and the CHIQ values are based on dissemination area (DA). The reason for this is because beginning with the 2001 census, the DA replaced the EA as the basic unit for geographical analyses used by Statistics Canada.
   
2. The SIQ method uses the Census definition of urban/rural. See the related information on Urban, Urban/Rural and Strictly Rural in the Income Quintiles Based on the 1996 Census concept for more information on how urban and rural areas are defined by Statistics Canada. The CHIQ method defines Winnipeg and Brandon as urban areas and considers everything else as rural.

Urban and Rural Income Quintile Considerations

Urban and Rural income quintiles are considered separately in MCHP research because of the difference in population density and income distribution in Manitoba. For current income quintile analysis, urban areas in Manitoba include Winnipeg and Brandon. All other areas are considered rural. The urban/rural distinction is made using municipal codes and where necessary, postal codes to define geographical areas.

Winnipeg is the largest city in Manitoba and the majority of the Manitoba population lives in Winnipeg. Brandon is also considered an urban area, but is much smaller than Winnipeg in size and population. The following population counts are reported by Statistics Canada from the 2016 Census data:

AREA	POPULATION COUNT
Manitoba	1,278,365 1
Winnipeg Census Metropolitan Area (CMA)	778,489 2
Winnipeg Census SubDivision (CSD)	705,244 3
Brandon Census Agglomeration (CA)	58,003 4
Brandon Census SubDivision (CSD)	48,859 5

Source: Statistics Canada Web Site: Census Profile, 2016 Census - accessed April 30, 2018.

¹ - search term "Manitoba" - select "Manitoba - Province, Population, 2016
² - search term "Winnipeg" - select "Census metropolitan areas - Winnipeg (CMA)"
³ - search term "Winnipeg" - select "Census subdivisions (municipalities) - Winnipeg (city)"
⁴ - search term "Brandon" - select "Census metropolitan areas / Census agglomerations - Brandon (CA)"
⁵ - search term "Brandon" - select "Census subdivisions (municipalities) - Brandon (city)"

Rural postal codes and dissemination areas (DA) typically cover a much larger geographical area than urban areas. For some variables, there is a concern that the larger rural areas are more heterogeneous; the area-level measure will be less correlated with the individual measure.

Income Quintile Ranges & Values

Information on income quintile ranges and values is currently available from two sources.


1. MCHP produces a document that provides income quintile information for each year from 1979 to 2022 using the CHIQ (Child Health Income Quintile) method. For each year and for each of the five rural and five urban income quintile rankings (levels), the minimum and maximum values are presented. This document is available at: Rural and Urban Income Quintiles for Manitoba - 1979 to 2022 (PDF).
   
   NOTE: This document contains two tables for each of the years 2008-2010. The reason for this is because of a change in methodology introduced in 2012 where several new postal codes related to Child and Family Services (CFS) offices were identified and removed from the analyses. These three tables reflect this change and are presented together on one page in the document. Information for these tables was taken from the 2006 Census and is presented in the document because values have been used in MCHP research. Subsequent years reflect this change in methodology. The second 2009 and 2010 tables presented, use data from the 2011 Census.
   
   Income quintile data for the earliest years (1979 to 1983) reported in the document are based on the 1981 Census household income values at the EA level. The 1976 Census did not collect household income information, therefore our range of years reported does not go back further than 1979. For more information on how this "early" information was developed, please read the Household Income from Census Information - Historical concept.
   
   
2. The second source, previously published information on Income Quintile Ranges and Values, is available in the MCHP concept: Household Income Value Ranges - 1984 to 1997. The information in this concept covers the years 1984-1997 inclusive, and is based on average household income from the 1996 Census and uses the Standard Income Quintile methodology employed at that time.

SAS Code and Formats

For our Data Analysts, SAS® formats for Income Quintiles are now available for each year from 1979 to 2022. Temporary formats for the years 2019 and 2020, based on the 2016 Census data are available for historic reasons, but formats for 2019 - 2022, based on the 2021 Census and the most recent Population and Postal Code Conversion File (PCCF) data are now available.

Information on the format names and details of the SAS macro code (_assign_iq) available in the MCHP SAS Macro Library are available internally through the LINKS section below: Income Quintile Formats and SAS Code - Updated for 2022 - INTERNAL - (internal access only).

A Variation on the Child Health Income Quintile (CHIQ) Methodology

In the Health Inequities in Manitoba: Is the Socioeconomic Gap in Health Widening or Narrowing Over Time? deliverable by Martens et al. (2010), income quintiles are used as a proxy measure of socioeconomic status. The analyses in this deliverable investigated census data indicators that are not readily available in the MCHP Data Repository, such as the percent of the population who are of aboriginal origin or the percent who are immigrants. Please read Chapter 2: A Description of the Neighbourhood Income Quintiles using Census Data (PDF) of the deliverable for a detailed description of the methodology used to generate income quintiles for this specific MCHP research.

The deliverable presents tables that characterizes and compares information from the 1986, 1996, and 2006 census years for each neighbourhood income quintile. For each income quintile, this includes the population size and mean (average) income, as well as the percent of Inuit, Metis, North American Indian, Aboriginal origin, Immigrant, unemployed, completed high school and private dwellings in need of major repairs. These tables can be viewed by clicking on the following links:


• Table 2.1: Neighbourhood Income Quintile Group Description Chart 1986 Census (PDF)
  
• Table 2.2: Neighbourhood Income Quintile Group Description Chart 1996 Census (PDF)
  
• Table 2.3: Neighbourhood Income Quintile Group Description Chart 2006 Census (PDF)

Important Notes

• Each person within an area is "attributed" the average household income of the area, so the income measure is not an individual income, but rather an area-level income measure.
  
• Work completed in July, 2015 by Shelley D. provides the availability of Income Quintile formats back to 1979, based on the CHIQ (Child Health Income Quintile) methodology. Subsequent years follow this same methodology.
  
• In May 2012, newly identified CFS postal codes have been moved into the income quintile not found group (NF). For more information on this group, see the Income Unknown (Income Not Found (NF)) glossary term.
  
• Income Quintile formats are updated on a regular basis, based on the availability of new Census data (every 5 years), updated PCCF files (annually), and updated (annually) Population files.
  
• When average household income values are available from new Census files, Income Quintile formats dating back two years from the current Census year are recreated. For example, when the 2021 Census data becomes available, new formats for 2019 and 2020 will be recreated using average household income data from the 2021 census file.