What does it mean to treat population growth as an underwriting input? It means converting a demographic statistic into a specific, checkable assumption about one property's revenue, at the geography where that property actually competes, and then testing what happens if the assumption turns out to be wrong. Used as a talking point, Sun Belt population growth is a reason to feel good about a purchase. Used as an input, it is four separate questions: which geography, which component of the growth, which vintage of the estimate, and by what mechanism a resident becomes rent.
Most marketing material answers none of the four. This guide works through each one against the figures the Census Bureau actually published, then covers the conversion rates that sit between a resident and a rent check and the discipline that keeps a growth assumption from becoming an unlabeled forecast. The regional case these figures are usually cited to support is set out in the guide to Texas commercial real estate investing and the Sun Belt.
Key Takeaways
- A population figure becomes an underwriting input only when a geography, a component of change, a vintage, and a specific revenue line are attached to it.
- Growth inside large metros is concentrated at the edges. Harris County accounted for roughly 38% of Houston's gain and Collin County roughly 35% of Dallas-Fort Worth's.
- Net international migration contributed a national rate of 3.70 per 1,000 in 2024 to 2025, more than twice the 1.52 rate from natural increase.
- That same component fell from 8.08 per 1,000 the prior year, so a growth assumption resting on it is resting on the fastest-moving of the three.
- Net domestic migration nets to zero nationally, because a move between U.S. counties subtracts where it leaves and adds where it arrives.
- Every figure here is a modeled estimate covering one twelve-month window, revised between vintages, which is why extrapolating it is a forecast rather than an assumption.
| Rate per 1,000 residents | |
|---|---|
| Natural increase | 1.5 |
| Net international migration | 3.7 |
| Net domestic migration | 0 |
Source: U.S. Census Bureau, Vintage 2025 Population Estimates.
What Turns a Growth Statistic Into an Underwriting Input
An input is a number that changes an output, which is a higher bar than a number that appears in a presentation. A population figure becomes an underwriting input only when four things are attached to it: the geography it describes, the component of change behind it, the vintage of the estimate, and the revenue line it is claimed to move.
Each attachment does different work. Geography decides whether the number describes the property's competitive area or a region containing it. Component decides whether the growth is the kind that persists. Vintage decides whether the figure can be cited and re-checked later. And the revenue line decides whether the statistic is doing anything at all, because a number that touches no line item in the model is a number that cannot be wrong.
That last point is the useful test. An assumption that cannot be falsified is not an assumption, it is a sentiment. Population growth used as a talking point is unfalsifiable by design: if the property performs, growth explains it, and if the property struggles, growth was never the specific claim. Writing down what would prove the input wrong converts it into something a reviewer can argue with.
It is also worth separating a supporting condition from a projected line. Demographic growth in a region is a supporting condition, meaning it raises the odds that a location remains viable over a long hold. A projected line is a number in a model, such as a rent increase or a renewal probability. Moving a supporting condition into a projected line without a stated conversion rate is the single most common error in this area, and the sections below are mostly about that conversion.
The Geography Problem: Metro, County, Trade Area
Population data is published at nested geographies, and a property competes at the smallest one. A metropolitan statistical area can span a dozen counties, a county can span a dozen trade areas, and growth inside those containers is uneven. Using a metro figure for a specific site quietly assumes the growth was distributed evenly, and it was not.
The Census Bureau's Vintage 2025 estimates make the concentration visible. For the year ending July 1, 2025, Houston led all metro areas with a gain of 126,720 residents and Dallas-Fort Worth followed with 123,557, ahead of Atlanta at 61,953, Phoenix at 59,065, Charlotte at 54,122, Austin at 53,796, and San Antonio at 38,402. At the county level, Harris County added 48,695 and Collin County 42,966, with Maricopa County at 35,411, Montgomery County, Texas at 30,011, and Wake County, North Carolina at 27,760.
Set the two lists against each other and the arithmetic is instructive. Harris County accounted for roughly 38% of the Houston metro's entire gain, and Collin County roughly 35% of Dallas-Fort Worth's, which means a single county in each case carried better than a third of a multi-county total. Every other county in those metros divided the remainder.
The Bureau described the pattern directly, noting that among some of the largest metro areas the fastest-growing counties tended to be on the outer edges, a pattern it called especially pronounced in Texas. Read as an underwriting instruction, that sentence says a mature inner-ring location can sit inside the second-fastest-growing metro in the country while its own trade area adds very little. The retail version of that geography, submarket by submarket, is worked through in the guide to the Dallas-Fort Worth retail market and the suburban build-out.
Components of Change: Where the Growth Came From
Components of change are the three sources any population figure comes from: natural increase, meaning births minus deaths, net domestic migration between U.S. areas, and net international migration. They behave differently and persist for different reasons, and a headline gain says nothing about which one produced it.
Natural Increase
The slowest-moving component, and the least sensitive to conditions in any given year. The Census Bureau reported a national natural increase rate of 1.52 per 1,000 residents for 2024 to 2025. It reflects the age structure of a population, which changes over decades rather than quarters.
Net Domestic Migration
Reported at a national rate of 0.00 per 1,000, which is arithmetic rather than news. A move from one U.S. county to another subtracts a person where it leaves and adds one where it arrives, so domestic migration redistributes the population without adding to it. A Sun Belt metro's domestic in-migration is another metro's loss, which makes it the component most dependent on conditions somewhere else.
Net International Migration
The largest contributor in the period and the most volatile. The Bureau reported a national rate of 3.70 per 1,000 for 2024 to 2025, more than double the natural increase rate, and down from 8.08 per 1,000 the prior year. As the comparison shows, one component supplied most of the growth and that same component fell by more than half in a single year.
The underwriting consequence is specific. A ten-year assumption built on a metro's recent gain is inheriting whatever mix produced it, and the three components carry different durability. The same components are reported below the national level, which is where the question becomes answerable for a particular market rather than rhetorical.
The Conversion Rates Nobody Writes Down
A resident does not pay rent. Four conversions sit between the two, and each has a rate that is almost never stated: residents to households, households to spending in a category, category spending to one store's sales, and store sales to a renewal decision at the end of the term. A growth percentage cannot be multiplied through a chain whose multipliers are unknown.
The first conversion is the one most often treated as one to one. Residents are not households, and the ratio between them moves with household size and age structure, so a metro adding residents does not add a proportional number of new households forming new shopping patterns. A gain concentrated in children within existing households is a different demand event from the same gain arriving as new household formations.
The second and third conversions narrow the funnel further. Households spend in categories rather than at addresses, and a category's spending is divided among every competing location within reach. A trade area can add households and send the incremental spending to a newer store, which is why competitive supply belongs in the analysis at exactly the point where population growth is usually allowed to stand alone.
The fourth conversion is the only one an owner touches, and it arrives once every ten to fifteen years. Renewal is a store-level decision made by people reading store-level numbers, which is why how comfortably a location's own sales cover its rent is more informative than any regional statistic. The method for reading that is set out in the guide to how to evaluate a net lease tenant and its rent coverage. Growth in the metro improves the odds on that last link. It does not set its value.
Estimates Are Estimates: Vintage, Revision, and Extrapolation
Every figure above is a modeled estimate rather than a count, published in a numbered vintage, revised in the next one, and covering a single twelve-month window. Treating a one-year measurement as a ten-year trend is the most common way a demographic input turns into an unlabeled forecast, and the label is the part that matters.
Three habits follow from that. The vintage and the reference period belong next to the number every time it is used, because a figure from one release will not match a figure from another and the difference is methodology rather than error. Estimates should be re-pulled rather than carried forward, since a county's gain can be revised between vintages. And any statement about what growth will do in future years should be written as a scenario, with the conditions it depends on named, rather than as an assumption sitting quietly in a cell.
The extrapolation trap is worth stating plainly. A model that only works if the metro keeps growing at the rate it grew last year contains a prediction dressed as an input. The drivers behind recent migration patterns, including employer relocations, remote work norms, and housing affordability, can change faster than a ten-year hold, and the volatility in the international migration component is a live illustration of how quickly one input can move.
None of this reprices an asset in the near term either. What sets the price at purchase is mostly tenant credit, remaining lease term, and the prevailing Treasury benchmark, which is the transmission covered in the guide to what a higher 10-year Treasury does to commercial real estate. Demographics support durability over a hold. They are not a premium collected at closing.
Writing the Assumption Down
Writing the assumption down is what separates an input from a talking point. Six steps do it, and all six are completed before an offer rather than after, which is the only point at which the answers are still useful.
- Name the geography and the vintage beside the number. County and place-level estimates for the few miles around the property, with the release they came from, rather than the metro aggregate.
- Classify the component. Whether the gain came mostly from natural increase, from domestic in-migration, or from international migration, and what each one depends on continuing.
- Keep growth out of the base case. A base case that holds demand flat and treats growth as upside produces a model that can be wrong about demographics without being wrong about the investment.
- Test the downside explicitly. What the location looks like if trade-area growth stops, and what it looks like if the growth arrives but a competing store captures it.
- Check the supply response. Announced centers, build-to-suit programs, and pad sites in the submarket, because demand met by new construction does not reach an existing owner as higher rent.
- Record the falsifier. The specific observation that would show the assumption was wrong, written where a reviewer will read it.
Investors reaching these markets through a pooled vehicle rather than by buying a building do a translated version of the same work, since the site selection belongs to the operator. The six steps become questions about process: which geography the growth data is pulled at, which vintage, whether the component mix is examined, and whether the competing pipeline is checked in each submarket. The specificity of those answers tends to be informative on its own.
Freedom Commercial Real Estate is a Dallas-based commercial real estate firm that publishes investor education, and this guide exists because a population statistic usually arrives as a single sentence with no geography, no vintage, and no stated effect on anything. A question about anything above, or a data source worth taking apart next, is welcome at info@freedomcre.net, and the guide to Texas commercial real estate linked earlier is the natural next read.
Frequently Asked Questions
Q: Which Sun Belt metros grew the most in population?
A: By numeric gain for the year ending July 1, 2025, the Census Bureau's Vintage 2025 estimates put Houston first at 126,720 residents added and Dallas-Fort Worth second at 123,557, with Atlanta at 61,953, Phoenix at 59,065, Charlotte at 54,122, Austin at 53,796, and San Antonio at 38,402. These are modeled estimates covering multi-county metro areas, so growth within each one is uneven.
Q: What are components of change, and why do they matter for underwriting?
A: They are the three sources of a population change: natural increase, meaning births minus deaths, net domestic migration between U.S. areas, and net international migration. They matter because durability differs. Natural increase moves slowly, domestic migration depends on conditions in the places people are leaving, and international migration was the largest national contributor in 2024 to 2025 at 3.70 per 1,000 after falling from 8.08 the prior year.
Q: Does population growth mean rents will rise?
A: Not by itself. Four unstated conversions sit between a resident and a rent check, and the last one, a tenant's renewal decision, turns on that store's own sales rather than on a regional figure. New construction also absorbs demand where land and entitlement allow it, so a submarket's supply pipeline is often the more decisive input of the two.
Q: How should population data be used in a real estate underwriting model?
A: As a supporting condition with a stated geography and vintage rather than as a projected line item. Common practice is to classify the component behind the gain, hold demand flat in the base case so growth reads as upside, test what the location looks like if growth stops or is captured by a competitor, and label any forward-looking statement as a scenario.
Sources
- U.S. Census Bureau, Vintage 2025 Population Estimates for Metro Areas and Counties
- U.S. Census Bureau, Population Estimates Program Glossary
Have a question about what you just read? Freedom Commercial Real Estate publishes these guides as investor education. Send the team your question, or keep reading below.
Get The Freedom Brief in your inbox
On the 15th of each month: the three guides most worth your time, the month's series, a stewardship reflection, and one sourced number. Education only, never a sales pitch. Unsubscribe any time.
This article is for informational purposes only and is not an offer to sell or a solicitation of an offer to buy any security, nor investment, tax, or legal advice. Any offering is made only through official offering documents to verified accredited investors. Past performance does not guarantee future results.
