Methodology for the ETI Urban Purchasing Power Profiles and Urban Markets Retail Sales Leakage/Surplus Reports

The University of Wisconsin-Milwaukee Employment and Training Institute provides comparison data on purchasing power, business activity, and workforce density for all residential ZIP codes and the 100 largest metro areas in the U.S. The profiles are designed to help cities, businesses, developers, and organizations assess the advantages of urban density for underserved city neighborhoods.

Each year, the U.S. Census Bureau conducts detailed analyses of the spending habits of residents by their household size, type, and income levels. These Consumer Expenditures Surveys are used by marketing firms to estimate expenditures of consumers and to assist retail companies in deciding where to locate and what populations to target for their consumer items and services. In most cases, marketing firms have based their data and recommendations on median household income of residents by geographic locations, while ignoring the advantages of urban density and the concentrated spending that takes place in city neighborhoods.

The purchasing power profiles developed by the Employment and Training Institute provide an easy-to-comprehend method to estimate purchasing power for common retail expenditures. Five steps are involved in the process. The ETI Purchasing Power Profiles presented here are based on analyses of 2002 and 2003 Consumer Expenditure Surveys (CEX) and the 2000 U.S. Census. Note: The reports have not been updated for more recent CES or Census data.

STEP #1: Using the CEX to determine what households buy.

The ETI Purchasing Power Profiles are based on spending patterns taken from the 2002 Consumer Expenditure Survey, utilizing survey responses from more than 30,000 interviews of households with complete income and expenditure responses. Additional data are drawn from the 2002 and 2003 CEX diary files, which includes patterns of spending by more than 22,000 respondents with complete income and expenditure data. The Purchasing Power Profiles focus on 16 categories of expenditures:

includes expenditures for food purchased at grocery stores and convenience stores, and food prepared at home for out-of-town trips.

includes expenditures for meals at restaurants, carry-out orders, food purchased on out-of-town trips, school lunches, and meals as pay.

includes expenditures clothing (suits, coats, sweaters, shirts, skirts, nightware, undergarments, hosiery, uniforms, costumes, etc.), accessories, footwear, material for making clothes, watches, jewelry, shoe repair, laundry and dry cleaning costs, and clothing storage.

includes expenditures for TVs, VCRs and video disc players; video cassettes, tapes and discs; video game hardware and software; cable and satellite service; repairs of TVs, radio and sound equipment; and rental of televisions.

includes expenditures for radios; tape recorders and players; sound components and component systems; records, CDs, audio tapes, and needles; record, tape, CD and video mail order clubs; musical instruments; accessories and other sound equipment; satellite dishes; and rental of above equipment.

includes expenditures for bathroom, bedroom, kitchen and dining room linens; curtains and draperies; slipcovers and decorative pillows; sewing materials for the home.

includes expenditures for mattresses and springs; sofas; living room tables and chairs; kitchen and dining room furniture; infants' furniture; outdoor furniture; wall units, cabinets and other occasional furniture.

includes expenditures for wall-to-wall carpeting (for renters and homeowners) and non-permanent floor coverings.

includes expenditures for dishwashers, garbage disposals, refrigerators, freezers, washing machines, clothes dryers, cooking stoves, microwave ovens, air conditioners; floor cleaning equipment, and sewing machines.

includes expenditures for china, dinnerware, flatware, glassware, serving pieces, small electric kitchen appliances, and portable heating and cooling equipment.

includes expenditures for computers, computer hardware, computer software and accessories, for nonbusiness use.

includes expenditures for window coverings, infants' equipment, outdoor equipment, clocks, lamps and lighting fixtures; other household decorative items; telephones and accessories; lawn and garden equipment; power tools; hand tools; plants and fresh flowers; closet and storage items; rental of furniture; and luggage.

includes expenditures for non- prescription drugs, non-prescription vitamins, eyeglasses and contact lenses, topicals and dressings, medical equipment for general use, supportive and convalescent medical equipment, and rental and repair of medical equipment.

includes expenditures for laundry and cleaning supplies, cleansing and toilet tissue, paper towels and napkins, miscellaneous household products, and lawn and garden supplies.

includes expenditures for hair care products, nonelectric articles for the hair, wigs and hairpieces, oral hygiene products and articles, shaving needs, cosmetics, perfume, bath preparation products, deodorants, feminine hygiene articles, and miscellaneous personal care items.

includes expenditures for paints; wallpapers; electrical supplies for heating and cooling equipment; materials for hard surface flooring, repair and replacement; materials and equipment for roof and gutters; materials for plastering, paneling, siding, windows, doors, screens, awnings; materials for patios, walks, fences, driveways, brick, masonry and stucco work; materials for landscaping maintenance; materials to finish basements, remodel rooms, or build patios, walks, etc.

Note: The ETI Purchasing Power Profiles do not include certain high-end purchases often included in national marketing firm reports (e.g., cars and boats), which are better identified through other data bases.

STEP #2: Determining expenditures by 5 household types and 5 income levels.

The Consumer Expenditure Survey provides data on spending by income levels and family types, which makes it possible to estimate expenditures within communities. For the ETI Purchasing Power Profiles, five types of households and five levels of income ranges by the same family types and income levels using 2000 U.S. Census data were considered in estimating expenditures for each of the 16 retail categories listed above. The five household types are:

  1. Families with children under age 18, with married parents.
  2. Families with children under age 18, with a single parent.
  3. Families with no children under age 18, with married heads of household.
  4. Families with no children under age 18, with a single head of household.
  5. Non-family households.

The annual household income ranges used are:

  1. $0 - $24,999
  2. $25,000 - $49,999
  3. $50,000 - $74,999
  4. $75,000 - $99,999
  5. $100,000 and above

In all, 25 expenditure estimates (i.e., 5 household types X 5 income levels) are calculated separately for each of 16 categories of expenditures.

The table below presents the estimates, based on CEX data, of annual expenditures for food purchased for the home by the 25 household/income cells. As shown in the table, expenditures for food at home by type of household increase only gradually as income doubles, triples, or even quadruples.

Annual Expenditures by Annual Household Income:
$0- $25,000- $50,000- $75,000- $100,000-
Type of Household $24,999 $49,999 $74,999 $99,999 and above
Families with children under age 18
     With married parents $4,809 $4,953 $5,317 $5,684 $6,893
     With single parent $3,685 $4,117 $5,126 $5,488 $6,141
Families with no children under age 18
     With married parents $3,430 $3,780 $4,190 $4,635 $5,327
     With single parent $3,159 $3,665 $4,108 $4,329 $5,065
Non-family households $1,944 $2,490 $2,856 $3,153 $3,613

The CEX data further shows that a large number of households in the U.S. are in the lowest income category and/or single parents -- groups often ignored or denigrated by marketing firm stereotypes.

Number of Households by Annual Household Income:
$0- $25,000- $50,000- $75,000- $100,000-
Type of Household $24,999 $49,999 $74,999 $99,999 and above
Families with children under age 18
     With married parents 4,893,474 2,922,042 1,076,752 382,633 326,974
     With single parent 1,907,308 1,729,770 771,438 317,394 171,986
Families with no children under age 18
     With married parents 3,006,793 5,325,310 5,705,498 3,875,143 4,406,655
     With single parent 5,107,187 6,548,635 4,903,928 3,209,307 3,885,130
Non-family households 18,174,910 7,900,311 3,143,023 1,086,162 1,103,594

STEP #3: Determining the number of household types in each income category in each neighborhood.

The 2000 U.S. Census data are used to obtain estimates of the number of households in each of the 25 cells identified above at the zipcode, census tract, and block group level. U.S. Census Bureau definitions of geographic units are used.

A ZCTA is a statistical geographic entity that approximates the delivery area for a U.S. Postal Service five-digit or three-digit ZIP Code. ZCTAs are aggregations of census blocks that have the same predominant ZIP Code associated with the residential mailing address in the U.S. Census Bureau's Master Address File. Three-digit ZCTA codes are applied to large contiguous areas for which the U.S. Census Bureau does not have five-digit ZIP Code information in its Master Address File. ZCTAs do not precisely depict ZIP Code delivery areas, and do not include all ZIP Codes used for mail delivery. (Source: U.S. Census Bureau) Only ZCTAs with residential populations are included in the ETI analyses.

A block group consists of all census blocks having the same first digit of their four-digit identifying numbers within a census tract. Block groups generally contain between 600 and 3,000 people, with an optimum size of 1,500 people. Block groups on American Indian reservations, off-reservation trust lands, and "special places" (i.e., correctional institutions, military installations, college campuses, worker's dormitories, nursing homes, and group homes) must contain a minimum of 300 people. (Source: U.S. Census Bureau)

Census tracts are small, relatively permanent statistical subdivisions of a county or statisically equivalent entity delineated by local participants as part of the U.S. Census Bureau's Participant Statistical Areas Program. The primary purpose of census tracts is to provide a stable set of geographic units for the presentation of decennial census data. Census tracts in the U.S. generally have between 1,500 and 8,000 people, with an optimum size of 4,000 people. Counties and statistically equivalent entities with fewer than 1,500 people have a single census tract. Census tracts on American Indian reservations, off-reservation trust lands, and "special places" (see above) must contain a minimum of 1,000 people. (Source: U.S. Census Bureau)

The PMSA consists of a large urbanized county or cluster of counties (cities and towns in New England) that demonstrate very strong internal economic and social links. The title of a PMSA may contain up to three place names (including the name of its largest central city and up to two additional central city names), sequenced in order of population from largest to smallest. (Metropolitan area definitions are determined by the Office of Mangement and Budget.)

STEP #4: Applying expenditure estimates against household/income data for the neighborhood.

Once the number of households are determined for the 25 cells (5 household types by 5 income levels), CEX expenditure patterns for each retail area are applied against the population in each of the 25 cells for each neighborhood. Separate calculations are made for each of the 16 retail categories, based on the findings of the 2002 and 2003 Consumer Expenditure Survey studies.

STEP #5: Calculating expenditures per square mile.

Emphasis on average household income by major marketing firms misses the significant spending by this large population of lower income families, and particularly the aggregate spending that occurs in dense urban neighborhoods. First, as seen above, families with lower incomes spend much higher percentages of their income on common retail purchases. Additionally, these families are often clustered in very dense neighborhoods while many upper income families reside in sparsely populated suburban or exurban areas.

The land area of each geographical unit is used to calculate density per square mile for the ETI Purchasing Power Profiles.

The area measurements of block groups, census tracts, and ZCTAs are calculated from the specific boundary recorded for each entity in the U.S. Census Bureau's geographic database, TIGER (the Topologically Integrated Geographic Encoding and Referencing). The U.S. Census Bureau provides area measurement data for both land area and total water area.

STEP #6: Calculating urban markets retail sales leakage or surplus.

Some neighborhoods are underserved by retail establishments and residents purchase many of their goods outside their community. Those census tracts where neighborhood retail sales fall below the estimated purchases of residents are said to have a retail sales leakage. That retail sales leakage is calculated by comparing the sales levels estimated from retail employment data with retail purchases from the purchasing power profiles. Retail sales surpluses occur in other tracts where retail sales estimated from retail employment data exceed local resident expenditures. These communities may have retail establishments attracting customers from outside the neighborhood, e.g. shoppers attracted to particular retail businesses, in-coming commuters, or stores serving metrowide markets.

Estimates are developed for all census tracts in the 100 largest metro areas in the U.S. to gauge retail sales activity in each neighborhood. To determine the extent to which existing retail businesses are capturing retail spending of local residents, consumer expenditures were estimated for 15 categories of consumer spending. All of the expenditure categories in the ETI Purchasing Power Profiles except for food-away-from-home (which is not in the NAICS retail sector) are included in the estimates for each census tract. This total is compared to estimates of retail sales derived from comparing employment in retail sales work by census tract with the total employment in retail sales for the metro area. Average retail sales per retail employee were calculated for each of the 100 largest MSAs by dividing the sum of the 15 categories of consumer spending estimated for the MSA by the number of retail workers in the MSA from census place-of-work tables. This average is then multiplied by the number of retail workers in each tract using place-of-work tables to get the estimated sales in the tract. Differences are reported as estimated retail sales "leakage" or "surplus."

The ETI Purchasing Power Profile methodology was developed by John Pawasarat, director of the UWM Employment and Training Institute; Dr. Frank Stetzer, senior information processing consultant for the UWM Information and Media Technologies Division; and Dr. Lois Quinn, senior scientist at the Employment and Training Institute.

In Milwaukee, the University of Wisconsin-Milwaukee Employment and Training Institute partnered with the City of Milwaukee Department of City Development to provide purchasing power analyses using current CEX data and income data more recent than the 2000 Census by using detailed ZIP code and block level data from the Wisconsin Department of Revenue on the most recent income tax year filings. This advanced methodology allowed the partners to adjust for changes in the economy and employment by neighborhood. These data offered comparable estimates for Wisconsin areas. At the same time, the nationwide statistics are available for comparisons with other metro areas throughout the U.S.

Free Purchasing Power Profiles, Retail Sales Leakage/Surplus Reports, Place-of-Work Drill Downs, and Workforce Diversity Drill Downs Available Online

How to Find Census Tracts in Your Community

You can locate the census tract for a specific address at the U.S. Census Bureau Factfinder Advanced Geography Search page using the GEOGRAPHY "address search" or "map" option.

For maps of census tracts in any community, go to the Click on "Census Tract Outline Maps 2000." Select your state, then county. Then select the PDF file for your county or select the first PDF file to locate the tracts for your part of the county.

To Top

Return to: | ETI Urban Purchasing Power/Workforce Density Page |
or the UWM Employment and Training Institute Home Page |

Page updated 2011
Employment and Training Institute
School of Continuing Education
University of Wisconsin-Milwaukee
Direct comments to