Can BLS and Census data pick your next coworking market?

What to take away

The pull order: QCEW, then OEWS, then LAUS

A US operator weighing a second site in Austin, Texas or Charlotte, North Carolina wants to know whether local demand can fill the desks. The Bureau of Labor Statistics tracks employment for each of those metros, and the Census Bureau counts the firms inside them.

The method below runs in a fixed order. QCEW county-industry tables first. OEWS metro occupation estimates second. LAUS metro rates third. Then a threshold test, then a field visit.

The promise of BLS employment series by metro is simple: where professional jobs grow, independent workers and small teams follow. Those workers rent desks. The logic holds until it does not.

The U.S. Bureau of Labor Statistics publishes employment, wage, and occupation data by metro and county. That makes it the first stop for any coworking demand analysis. The agency's own overview of BLS statistics explains the range of series available for market work.

Start with the series that track industries and occupations common among coworking members: professional services, information, finance, and management. If those counts rise, membership demand usually follows within a few quarters.

But labor data measures payroll jobs. Many coworking members are self-employed, contract, or remote employees of firms based elsewhere. Those people appear in labor data only faintly, if at all.

That gap is the central problem. A metro can show flat payroll employment while its freelance and remote-worker population grows fast. The data screen misses the very people who buy hot desks.

So the forecasting question has a two-part answer. Labor data can rank metros by professional density and momentum. It cannot confirm that a specific neighborhood, building, or price point will work. That confirmation comes from field checks and local knowledge.

For a wider view of where coworking demand is heading, see the coworking space business insurance line items that shape operator costs.

What QCEW and OEWS reveal about a metro's professional density

The Quarterly Census of Employment and Wages is the workhorse. It counts employment and wages by county and industry, covering most U.S. jobs. The QCEW program is the place to pull those county-industry tables.

QCEW data is granular. You can compare professional-services employment in two adjacent counties and see which one added jobs last year. That is useful when two suburbs look identical on a map.

OEWS, the Occupational Employment and Wage Statistics program, adds the occupation layer. It tells you how many management, computer, and business-operation jobs a metro has, and what they pay. The BLS occupation data overview is the entry point for that data.

High concentrations of computer and mathematical occupations often signal a strong coworking market. Those workers change jobs often, freelance on the side, and value flexible desks. The same is true for business and financial operations occupations.

Use QCEW and OEWS together. QCEW shows which industries are growing by county. OEWS shows which occupations dominate. A county adding professional-services jobs that are mostly accountants will behave differently from one adding software developers.

Here is the screen, with numbers attached. Pull the QCEW annual averages for NAICS sector 54, professional, scientific, and technical services, by county and industry, for the last three years. Austin-Round Rock-Georgetown is area code 12420 in those files, and Charlotte-Concord-Gastonia is 16740.

Then pull the May OEWS metro estimates for occupation group 15-0000, computer and mathematical, and 13-0000, business and financial operations. Both metros are large enough that OEWS publishes these counts rather than suppressing them.

A metro passes the first test when NAICS 54 employment is up at least 5 percent over three years, and those two occupation groups are 10 percent or more of total metro employment. Write both cutoffs down before you pull anything.

The screen removes metros with shrinking professional bases and keeps the ones worth a closer look. The next step is checking whether the growth sits in occupations that actually rent desks.

The industry data overview helps here. It groups data by sector so you can size demand from industries such as information, finance, and professional services without pulling every table separately.

For a ranked list of metros that clear a similar screen, see the markets to consider.

Where Census business patterns fall short for coworking demand

Census business patterns, especially the County Business Patterns and Nonemployer Statistics programs, add two things labor data lacks: establishment counts and self-employment counts. Both are published by county and NAICS industry.

Establishment counts matter. A metro with many small professional-services firms has more potential member companies than one with a few large employers. Small firms rent coworking space; large firms build out their own floors.

Nonemployer Statistics matter even more. They count businesses with no paid employees, which includes many freelancers and solo consultants. Those are core coworking members. The U.S. Census Bureau publishes these by county and industry.

Still, Census business patterns fall short in predictable ways. They lag. The most recent nonemployer data often trails the current market by two years or more. A metro can change a lot in two years.

They also miss remote employees. A software engineer living in Boise and working for a San Francisco firm appears in neither the Boise payroll data nor the Boise nonemployer counts. Yet that person may buy a coworking membership.

Industry codes blur. A one-person design studio and a large ad agency can share a NAICS code. Establishment counts do not tell you which firms are growing, hiring, or likely to need desks.

Geography blurs too. County lines rarely match how people move through a city. A coworking space near a county border may draw members from two counties with different data profiles.

Use Census business patterns as a breadth check, not a depth check. They tell you how many small firms and solo operators exist in an area. They do not tell you whether those operators want coworking space, can afford it, or will stay.

A hybrid data-plus-field test solves this. Data narrows the map. Field work answers the questions data cannot.

Testing a market with LAUS and regional BLS series

LAUS, the Local Area Unemployment Statistics program, tracks unemployment rates for states, metros, and counties. It is monthly, which makes it the timeliest labor signal for a coworking market. In the series ID, measure code 03 is the unemployment rate.

LAUS unemployment trends are easy to misread. A rising unemployment rate can mean layoffs, which hurt membership demand. It can also mean more people are entering the labor force, which can eventually help.

The useful signal is direction and duration. A metro where unemployment has fallen for six straight months is adding workers. A metro where it has risen for a year is losing them. Coworking demand follows that trend with a lag.

Pair LAUS with regional BLS series for the metro. Employment by industry, average weekly wages, and job openings all come from the same family of data, published for most large metros.

Here is a worked example. Suppose you are comparing the Austin-Round Rock metro in Texas with Charlotte-Concord-Gastonia in North Carolina. Austin's professional base leans on software and state government. Charlotte's leans on banking and finance.

Pull both metro rates for the same month from the BLS Metropolitan Area Employment and Unemployment release. It prints metro rates and carries a release date you can cite. Read the 12-month change, not the level.

A metro can carry the higher headline rate of the two while adding professional-services jobs faster. Momentum matters more than the level for a lease signed three years out.

Now add demographic data. The BLS demographic data overview covers age, race, and other characteristics of the workforce. A metro with a growing young professional population has more potential members than one with an aging workforce.

Demographics also shape product mix. A metro heavy in young freelancers may want hot desks and event space. A metro heavy in established consultants may want private offices and meeting rooms.

Use LAUS for timing, regional BLS series for direction, and demographic data for fit. Together they narrow the field to two or three candidates worth visiting.

Limits of data-driven coworking space operations decisions

The biggest limit is lag. QCEW data arrives months after the quarter ends. Census business patterns arrive years after the fact. LAUS is monthly but revised. By the time a trend is visible in the data, the market may have moved.

Gig and remote work hide. A growing share of the workforce is not captured cleanly in payroll or nonemployer data. Coworking demand analysis that relies only on official counts will undercount the fastest-growing segment.

Local shocks do not show up in advance. A major employer announces layoffs, a university expands, a transit line opens. Data reflects these events after they happen. Operators who wait for the data will be late.

Data cannot measure culture. Some cities have a strong independent-work culture and a weak coffee-shop scene. Others have the opposite. The numbers look similar; the membership demand does not.

Competition is invisible in the data. In a US metro like Chicago or Seattle, national operators and regional chains can add several floors within a few blocks in a single year, and no BLS table records that supply. Supply matters as much as demand.

Costs are invisible too. No BLS table carries the asking rent on the floor you want. That number comes from a dated broker report, such as CBRE's quarterly US office figures or Cushman and Wakefield's MarketBeat, and you cite it with the quarter it covers.

Finally, data cannot tell you whether a specific building works. Ceiling heights, natural light, transit access, and parking decide whether members renew. That is a field question.

The limits do not make data useless. They define its role. Data is a screen, not a verdict. It tells you where to look, not what to sign.

For the operational questions that data cannot answer, see the site selection explained.

A practical hybrid: data screen plus on-the-ground checks

A hybrid data-plus-field test uses data to rank metros, then uses field work to confirm or reject the ranking. The sequence matters. Data first, field second.

The table below writes out the same screen for the two metros, with the file behind each row and the cutoff that counts as passing:

Those cells are queries, not quotes. QCEW, OEWS, and LAUS get revised and re-released, so a percentage copied from an article goes stale before a lease is signed. Pull the rows the week you need them, and cite the vintage you pulled.

Now the field checks. Visit both metros. Spend a day in each. Work from a local coffee shop in the morning and a competitor's coworking space in the afternoon.

Checklist for the field visit:

Score each field item the same way you scored the data. If the field score contradicts the data score, trust the field. Data lags; the street does not.

A final check is regulatory and financial. The IRS publishes tax rules for business deductions that affect how members and operators structure expenses. The SBA offers loans and counseling for small-business operators, including coworking space owners. OSHA sets workplace safety rules that apply to shared spaces.

Local building codes, often based on the International Building Code, govern build-outs and occupancy. The National Coworking Association publishes best practices for operators.

None of these sources forecast demand. They shape the cost and risk side of the decision, which matters just as much.

A hybrid test does not guarantee success. It reduces the chance of a costly mistake. That is the realistic goal of coworking market data.

For a broader view of expansion questions, see the expansion and market guide. For comparing specific sites, see the locations worth comparing.

Common questions