A three-step workflow using the Multigenerational Household dimension, modeled from ACS 1-Year PUMS.
Every growth plan begins with a market size. How many households are there, where are they, how do they live, and how many of them can your institution serve. The discipline sounds simple until the segment you are sizing has no agreed definition, and multigenerational living is exactly that case.
Depending on which households you count, the U.S. multigenerational market is 5.1 million households or 26.7 million households. Both figures are correct. They answer different questions, and a growth plan built on the wrong one will misjudge the opportunity by a factor of five.
This is a workflow for sizing that market properly in CensusVision, using the Multigenerational Household dimension modeled from ACS 1-Year Public Use Microdata. Establish the universe, segment it, then bring it down to the markets you actually lend in.
Roughly 26.7 million U.S. households, about one in five, contain two or more adult generations under one roof. For a mortgage executive, that concentration touches several decisions at once.
None of these decisions can be made from a national average. They require the segment sized, segmented, and localized, which is what the workflow below produces.
Open the Census Filters sheet with no selections applied. The 2024 base is approximately 132.7 million households. Every share you calculate later comes back to this denominator, so anchor it first.
Then open the Multigenerational Household filter. It carries seven categories, and the sizing decision is really a decision about which ones to include:
Three defensible definitions fall out of this:
State which definition you are using whenever you publish a number. The 5x gap between strict and broad is where most confusion about market size comes from.
With your definition selected, apply the Ethnicity filter (Hispanic). The separate Hispanic Origin filter carries the disaggregated detail, Mexican, Salvadoran, Cuban, and more, for when the analysis needs to go below the pan-Hispanic total. One filter click produces two different measures, and keeping them straight is the difference between describing the market and understanding it.
Composition is a group's multigenerational households divided by all multigenerational households. It answers the question: who is the multigen market? Of the 26.7 million broad multigen universe, 6.5 million households are Hispanic, or 24.3%, well above the 15.2% Hispanic share of all households. Under the strict definition the concentration is sharper still: Hispanic households account for 28.9% of all 3+ generation households. Hispanic household count also grew 10.6% from 2021 to 2024, so this segment is expanding as it concentrates.
Propensity is a group's multigenerational households divided by all households in that group. It answers a different question: how does this group live? In CensusVision, clear the multigen selection, filter to Hispanic households to establish that sub-universe of 20.2 million, then re-apply the multigen selection. 32.1% of Hispanic households live multigenerationally under the broad definition, against 20.1% of all households and 16.1% of non-Hispanic White households. Hispanic propensity runs at twice the non-Hispanic White rate. Under the strict definition, the Hispanic share is 7.3% against 3.9% overall.
The two measures carry different jobs. Composition sizes the market and allocates outreach: it tells you that one in four broad-multigen households is Hispanic. Propensity explains behavior and drives product: it tells you that a Hispanic household you meet is far more likely to bring pooled adult incomes, potential co-borrowers, and demand for larger floor plans and next-generation suites. A group can dominate composition simply by being large. Elevated propensity is what signals a structurally different way of living, and the Hispanic segment is notable because it leads on both measures at once.
The same two-measure approach works for other Person Filters: Race, Citizenship, Veteran Status, or Hispanic Origin for the disaggregated view. Within a multigen segment, Relationship to Householder sets the direction of the arrangement, since a Father or mother present signals a parent who moved into an adult child's household, while an adult son or daughter signals a grown child in the parents' home, and Age Range refines life stage from there. Propensity comparisons across groups are where the underwriting and product implications live.
Multigenerational shares vary widely by market, and geography is where CensusVision separates from published tables. With your definition and segment held constant, work down the Geo Filters:
Finish with OwnVsRent and House Value Range from the Household Filters to convert the count into a mortgage question: how many multigen households own, at what value points, and where. The tenure cross-cut delivers a finding many lenders would not expect. Multigenerational households own at a premium in every group: 73.8% of broad-multigen households own against 65.3% of all households, 83.7% against 73.3% among non-Hispanic White households, and 61.4% against 50.9% among Hispanic households. Multigenerational formation functions as an ownership pathway, with pooled adult earners converting to homeownership across the board, and the Hispanic segment pairs that pathway with the highest propensity and fastest growth.
Keep the measure on Households when sizing the market. When you apply a person-level filter such as Ethnicity or Age Range to a household count, the selection includes every household containing at least one person meeting the condition. That is usually the intended behavior, and it is worth knowing precisely what the number represents before it goes in a deck.
National figures set the baseline. Decisions happen locally, and both measures change when you get there. A metro can hold a modest share of your footprint's multigen households while its propensity runs far above the national 20.1%, and those two facts point to different moves. Composition tells you where the multigen households are, so it shapes branch placement, loan officer coverage, and partnership outreach. Propensity tells you how households in a market live, so it shapes co-borrower and household-income guidelines, product design, and the conversation your originators are prepared to have.
Start a complimentary trial and run this workflow for your states, metros, and PUMAs, so that your growth plan reflects how the households in your market actually live rather than a national average.
Methodology. This analysis uses the American Community Survey 1-Year Public Use Microdata Sample (ACS PUMS, 2024), as modeled in CensusVision by Polygon Research. Multigenerational classifications follow Census Bureau and IPUMS definitions. Household counts reflect occupied housing units. Hispanic and non-Hispanic White household figures use the Ethnicity and Race filters applied to household counts; the Hispanic Origin filter provides disaggregated detail below the pan-Hispanic total. Composition divides a group's multigenerational households by all multigenerational households. Propensity divides a group's multigenerational households by all households in that group. Person-level filters applied to household counts select households containing at least one person meeting the condition. Growth rates compare 2021 and 2024 ACS 1-Year PUMS.
Market sizing starts with a definition. Multigenerational living ranges from 5.1 million to 26.7 million households depending on how you count, and the choice changes product, underwriting, and growth strategy.
A three-step workflow using the Multigenerational Household dimension, modeled from ACS 1-Year PUMS.
Every growth plan begins with a market size. How many households are there, where are they, how do they live, and how many of them can your institution serve. The discipline sounds simple until the segment you are sizing has no agreed definition, and multigenerational living is exactly that case.
Depending on which households you count, the U.S. multigenerational market is 5.1 million households or 26.7 million households. Both figures are correct. They answer different questions, and a growth plan built on the wrong one will misjudge the opportunity by a factor of five.
This is a workflow for sizing that market properly in CensusVision, using the Multigenerational Household dimension modeled from ACS 1-Year Public Use Microdata. Establish the universe, segment it, then bring it down to the markets you actually lend in.
Roughly 26.7 million U.S. households, about one in five, contain two or more adult generations under one roof. For a mortgage executive, that concentration touches several decisions at once.
None of these decisions can be made from a national average. They require the segment sized, segmented, and localized, which is what the workflow below produces.
Open the Census Filters sheet with no selections applied. The 2024 base is approximately 132.7 million households. Every share you calculate later comes back to this denominator, so anchor it first.
Then open the Multigenerational Household filter. It carries seven categories, and the sizing decision is really a decision about which ones to include:
Three defensible definitions fall out of this:
State which definition you are using whenever you publish a number. The 5x gap between strict and broad is where most confusion about market size comes from.
With your definition selected, apply the Ethnicity filter (Hispanic). The separate Hispanic Origin filter carries the disaggregated detail, Mexican, Salvadoran, Cuban, and more, for when the analysis needs to go below the pan-Hispanic total. One filter click produces two different measures, and keeping them straight is the difference between describing the market and understanding it.
Composition is a group's multigenerational households divided by all multigenerational households. It answers the question: who is the multigen market? Of the 26.7 million broad multigen universe, 6.5 million households are Hispanic, or 24.3%, well above the 15.2% Hispanic share of all households. Under the strict definition the concentration is sharper still: Hispanic households account for 28.9% of all 3+ generation households. Hispanic household count also grew 10.6% from 2021 to 2024, so this segment is expanding as it concentrates.
Propensity is a group's multigenerational households divided by all households in that group. It answers a different question: how does this group live? In CensusVision, clear the multigen selection, filter to Hispanic households to establish that sub-universe of 20.2 million, then re-apply the multigen selection. 32.1% of Hispanic households live multigenerationally under the broad definition, against 20.1% of all households and 16.1% of non-Hispanic White households. Hispanic propensity runs at twice the non-Hispanic White rate. Under the strict definition, the Hispanic share is 7.3% against 3.9% overall.
The two measures carry different jobs. Composition sizes the market and allocates outreach: it tells you that one in four broad-multigen households is Hispanic. Propensity explains behavior and drives product: it tells you that a Hispanic household you meet is far more likely to bring pooled adult incomes, potential co-borrowers, and demand for larger floor plans and next-generation suites. A group can dominate composition simply by being large. Elevated propensity is what signals a structurally different way of living, and the Hispanic segment is notable because it leads on both measures at once.
The same two-measure approach works for other Person Filters: Race, Citizenship, Veteran Status, or Hispanic Origin for the disaggregated view. Within a multigen segment, Relationship to Householder sets the direction of the arrangement, since a Father or mother present signals a parent who moved into an adult child's household, while an adult son or daughter signals a grown child in the parents' home, and Age Range refines life stage from there. Propensity comparisons across groups are where the underwriting and product implications live.
Multigenerational shares vary widely by market, and geography is where CensusVision separates from published tables. With your definition and segment held constant, work down the Geo Filters:
Finish with OwnVsRent and House Value Range from the Household Filters to convert the count into a mortgage question: how many multigen households own, at what value points, and where. The tenure cross-cut delivers a finding many lenders would not expect. Multigenerational households own at a premium in every group: 73.8% of broad-multigen households own against 65.3% of all households, 83.7% against 73.3% among non-Hispanic White households, and 61.4% against 50.9% among Hispanic households. Multigenerational formation functions as an ownership pathway, with pooled adult earners converting to homeownership across the board, and the Hispanic segment pairs that pathway with the highest propensity and fastest growth.
Keep the measure on Households when sizing the market. When you apply a person-level filter such as Ethnicity or Age Range to a household count, the selection includes every household containing at least one person meeting the condition. That is usually the intended behavior, and it is worth knowing precisely what the number represents before it goes in a deck.
National figures set the baseline. Decisions happen locally, and both measures change when you get there. A metro can hold a modest share of your footprint's multigen households while its propensity runs far above the national 20.1%, and those two facts point to different moves. Composition tells you where the multigen households are, so it shapes branch placement, loan officer coverage, and partnership outreach. Propensity tells you how households in a market live, so it shapes co-borrower and household-income guidelines, product design, and the conversation your originators are prepared to have.
Start a complimentary trial and run this workflow for your states, metros, and PUMAs, so that your growth plan reflects how the households in your market actually live rather than a national average.
Methodology. This analysis uses the American Community Survey 1-Year Public Use Microdata Sample (ACS PUMS, 2024), as modeled in CensusVision by Polygon Research. Multigenerational classifications follow Census Bureau and IPUMS definitions. Household counts reflect occupied housing units. Hispanic and non-Hispanic White household figures use the Ethnicity and Race filters applied to household counts; the Hispanic Origin filter provides disaggregated detail below the pan-Hispanic total. Composition divides a group's multigenerational households by all multigenerational households. Propensity divides a group's multigenerational households by all households in that group. Person-level filters applied to household counts select households containing at least one person meeting the condition. Growth rates compare 2021 and 2024 ACS 1-Year PUMS.
Under the strict Census Bureau definition of three or more generations, about 5.1 million US households were multigenerational in 2024, based on ACS 1-Year PUMS modeled in CensusVision. A broader definition including all shared adult generations counts roughly 26.7 million households.
The Census Bureau counts households with three or more generations. Broader definitions used in housing analysis add grandparent-grandchild households with no middle generation and households where adult children live with parents or in-laws. The definition chosen changes the market size by a factor of five. Two measures then describe any segment: composition, the group's share of all multigenerational households, and propensity, the multigenerational share within the group.
In 2024, 32.1% of Hispanic households lived in a shared-adult-generation arrangement, against 20.1% of all households, based on ACS 1-Year PUMS modeled in CensusVision. Under the strict 3+ generation definition, the shares are 7.3% for Hispanic households and 3.9% overall. Hispanic households account for 28.9% of all 3+ generation households.
CensusVision models the American Community Survey 1-Year Public Use Microdata Sample (ACS PUMS), applying Census Bureau and IPUMS multigenerational classifications. Estimates are available down to the PUMA level, below what standard published ACS tables provide at this cross-cut depth.