This week's Podcast/YouTube show
Guest: Erica Groshen, Labor Economist and Former Commissioner of the U.S. Bureau of Labor Statistics
Trust and the Bureau of Labor Statistics
Former BLS Commissioner Erica Groshen’s message is disarmingly simple: the monthly employment report remains the most important economic release in the world because it anchors business-cycle forecasts and, by extension, the cost of capital that underwrites commercial real estate.
When sponsors debate exit cap rates, debt availability, and absorption, they are, whether they realize it or not, running on the operating system that these statistics provide.
Watch the full conversation here >>
CES, CPS, QCEW - What Matters for Investors
Groshen distinguishes between the three headline surveys behind the employment data:
- CES (Current Employment Statistics): a monthly sample of employers used to estimate total nonfarm jobs in near-real time.
- CPS (Current Population Survey): a monthly survey of households used to estimate unemployment and labor force metrics.
- QCEW (Quarterly Census of Employment and Wages): administrative data from unemployment insurance (UI) wage filings employers provide to the stats that covers roughly 95% of jobs; slower, but more complete, and used to benchmark CES annually.
For real estate professionals, the practical point is timing versus completeness. CES gives a timely read on cyclical turning points that influence debt markets long before QCEW arrives. QCEW later “true-ups” the level, but lenders and investors have already made decisions using the flow of monthly estimates. The implication: treat first prints as directional, not definitive.
About That Recent 911,000 “Miss” – Perspective Over Drama
The recent headline controversy: a 911,000 downward benchmark revision to previously reported job gains (March 2024 to March 2025). Groshen notes that, in the context of a 150+ million job base, that change amounts to roughly 0.6% of total employment, even if it is a larger share of the year’s cumulative growth.
Translation for sponsors: this is an adjustment simply showing that the magnitude of growth was slower than first reported, not evidence that the trend itself was wrong.
Why did it happen? Two structural forces Groshen highlights:
- Nonresponse/selection effects: establishments that declined to participate differed from those that did with this cycle skewing toward slower job creation among nonrespondents.
- Misreporting during the year: some employers over-stated payrolls in the monthly survey relative to later UI filings, e.g., by accidentally including contractors or making simple accounting errors.
Both are known issues; both are exactly what annual benchmarking is designed to detect and correct. For CRE, the takeaway isn’t to dismiss the monthly numbers but to incorporate revision risk into macro assumptions and sensitivity cases.
Precision, Not Certainty: Read the Error Bars
A point many market takes ignore: the standard error (something we would expect from normal sampling variability) on the monthly change in payrolls is over 100,000, and the BLS literally labels first prints as preliminary – which most consumers promptly ignore.
The right investor behavior is to weight the direction and the sequence of data, not any single datapoint. Ask: is breadth improving or narrowing by industry and region? Are hours and wages confirming? Is the unemployment rate moving with participation? These cross-checks matter more than fixating on one number.
Declining Response Rates: Why Granularity Is Getting Harder
Groshen underlines a secular headwind: response rates for surveys, public and private, have been falling for 30–40 years. In the payroll survey, the agreement to participate (initiation) rate has dropped from ~70% to ~35% in the past decade.At the national aggregate you still get robust signals, but granularity suffers by industry and local geography.
That matters for CRE because sectoral detail (e.g., retail trade, logistics, professional services) often drives leasing pipelines and tenant credit views. Underwrite with humility: accept that some micro-reads are noisier than they used to be and lean on corroborating indicators (job postings, hours, tax receipts) when possible.
Modernization: What “Better” Could Look Like
International peers show potential paths forward. Australia, for example, feeds payroll tax remittances, sanitized for privacy, directly into the statistical agency for near real-time reads. The U.S. could pursue public-private rails with payroll processors or common data standards to reduce burden and raise quality.
For investors, modernization would mean smaller, more frequent revisions and earlier visibility into inflection points, benefits that directly reduce underwriting uncertainty. The constraint, Groshen notes, is funding: the BLS has plans on the shelf but lacks sustained appropriations to execute them at scale.
Trust as an Economic Input
The most consequential section of the conversation is not technical, it’s institutional. Data systems only work if users trust they are objective and accurate.
Political attacks on statistical agencies erode usage, depress participation, and create a downward spiral in quality. As trust declines, firms stop responding; as data degrade, risk premia rise.
For CRE, that flows through into higher borrowing costs, wider exit cap assumptions, and slower investment. A world that flies “closer to blind,” as Groshen puts it, is a world with a higher discount rate.
What If the Fed’s Independence Falters?
Groshen’s warning is crisp: if monetary policy is set in the White House and data are bent to fit, you may initially get a short “sugar high” – lower rates, a burst of activity – followed by stagflation, rising distrust, and prolonged instability.
History offers examples (e.g., Argentina, Turkey, Greece). For sponsors, that scenario argues for
(1) conservative debt structures that survive both higher inflation and growth slowdowns,
(2) lease strategies that protect real rents, and
(3) option value in business plans (e.g., phasing, refi optionality, rate-cap hedging) that can flex across regime shifts.
Practical Reading List for Underwriting Meetings
When your investment committee or lender meeting turns to “the macro,” here’s a disciplined way to use BLS outputs:
- Look at 3–6 months of CES changes plus revisions; treat single-month spikes with skepticism.
- Cross-check with CPS (unemployment, participation) and average weekly hours for labor-demand confirmation.
- Track industry composition: which sectors are adding or shedding? Map to your rent roll and pipeline.
- Acknowledge revision risk in your model: run sensitivities for “growth was 0.5–0.7% lower” and see what breaks.
- Remember timing: QCEW will reconcile levels; don’t wait for perfection to make decisions, but don’t pretend prelims are final.
In short: the data are good enough to guide action, but not so perfect that you can skip prudence.
***
Erica’s ‘sugar rush’ comment resonates with my own sense of what is going to be happening in the coming months and years.
As confidence in the institutions that have made America great become increasingly eroded and politicized, expect a period of exuberance followed by a severe economic hangover.
Now is the time to be buying, investing, and staying focused on prudent underwriting, stress tested for long term viability and the inevitability of a severe downturn during the lifecycle - because when trust in institutions weakens, volatility rises and real estate always pays the price.
Watch/listen to full episode here.
Please join the conversation on LinkedIn, here.
Or feel free to reply to this email with any thoughts or comments.