The Economic
Systems Monitor
Tracking how pressure moves through the U.S. economy, what absorbs it, and whether the systems meant to respond are restoring function—or moving the burden somewhere else.
Where pressure
is moving now.
Individual indicators matter. But the central question here is relational: where stress originates, where it travels, what absorbs it, and whether the response restores function or transfers the burden.
What changed
most recently.
What happens
next?
DFTR uses the present configuration of relationships to predict how pressure will move next. Each forecast is issued in advance, dated, and retained so its accuracy can be evaluated.
Loading the prospective forecast…
What would
have to
be true?
The prediction is not evaluated by whether every new observation looks bad. It is evaluated against conditions established before the evidence arrived.
Does pressure begin to travel?
Stress that was previously contained begins appearing in adjacent systems, producing increasingly connected forms of strain.
OBSERVATIONS SUPPORTING THIS CONDITION SINCE BASELINECan correction restore function?
Systems may receive feedback and make bounded corrections while broader relational repair remains inadequate.
OBSERVATIONS SUPPORTING THIS CONDITION SINCE BASELINEWhat does stability begin to cost?
Repair becomes more expensive, extraordinary, or internally conflicting as institutions respond to pressure across multiple domains.
OBSERVATIONS SUPPORTING THIS CONDITION SINCE BASELINEWhat would weaken the prediction?
Sustained normalization, restored buffers, uncoupling between domains, and durable repair count against the prospective claim.
EVIDENCE OF NORMALIZATION SINCE BASELINE
What would
count as
the economy
collapsing?
P8 is the strongest claim in the prospective prediction: not ordinary economic weakness, a recessionary signal, or a period of market volatility, but a qualifying systemic disruption in which stress is no longer being adequately contained by the economy's ordinary buffers and repair mechanisms.
More than deterioration.
Rising rates, weaker hiring, household strain, falling confidence, credit stress, or declining investment can all matter without constituting systemic collapse on their own.
More than one failing domain.
The threshold concerns propagation: disruption moving across economically important relationships rather than remaining a contained problem inside a single market, institution, or sector.
Buffers and repair have to matter.
The critical question is whether households, firms, markets, and public institutions can continue absorbing pressure and restoring function—or whether those mechanisms become insufficient as stress compounds.
Loading published evidence…
Where things stand.
Loading the latest published evidence and system relationships…
Evidence snapshot based on approved observations. A formal change to the P8 threshold requires its own assessment.
Each domain shows its latest published observation and the balance of evidence over the past 30 days. Select one to see its supporting records below.
Loading domains…
What the
numbers miss.
An economy can show improvement in a headline statistic while people are still losing ground. Finding a job doesn't necessarily mean recovering the income, hours, or housing security someone had before.
These are limits on what an indicator can establish, not evidence that every reassuring number is wrong. The monitor tracks both deterioration and genuine recovery.
Are households and businesses recovering their ability to function, or are they continuing under conditions that are harder to sustain?
01 / LABORUnemployment isn't earnings recovery.
What it measures
The share of the labor force without work who meet the survey's unemployment criteria, including recent active job search or temporary layoff.
What it misses
People who want work but aren't actively searching, reduced hours, and workers who return to jobs paying much less than their previous jobs. Being employed doesn't establish that prior earning capacity has been restored.
What we check alongside it
Employment transitions, people outside the labor force who want a job, involuntary part-time work, displaced-worker outcomes, and earnings recovery.
BLS household survey ↗02 / SURVEY COVERAGESome hardship isn't captured by household surveys.
What it measures
The Current Population Survey estimates conditions among the civilian noninstitutional population covered by its household-based sample.
What it misses
People living unsheltered are generally outside its housing-unit sampling frame. Nonresponse can also affect estimates. These limitations don't, on their own, prove that the unemployment rate is understated or that any bias has grown since the baseline.
What we check alongside it
Survey response and coverage research, housing instability, homelessness counts, and administrative records where comparable.
Census CPS methodology ↗03 / HOUSING & WORKING HOMELESSNESSEmployment doesn't guarantee housing security.
What conventional measures miss
Employment statistics measure work, not whether wages cover housing. Mortgage delinquency excludes renters and people without mortgages; conventional household surveys and homelessness counts also have coverage limits. Track rent burdens, eviction filings, employment among people experiencing homelessness, vehicle dwelling, and exits into stable housing alongside labor-market data.
HUD homelessness reporting ↗What the evidence shows
A September 2026 California Policy Lab study linked homelessness-service and wage records in Los Angeles County. Among service enrollees studied, 80% had previously worked in California and 21% had formal employment at enrollment. Among those with earnings two years later, average annual earnings were $18,468. The study covers people who enrolled in services in 2019–2022, not a new 2026 national employment rate.
California Policy Lab research (September 2026) ↗People living in vehicles
King County researchers estimated approximately 3,967 people living in cars, vans, and RVs on one night in 2026, and about 5,460 over a year. Vehicle residents may be difficult to locate through conventional counts; these local estimates cannot be extrapolated into a national trend.
University of Washington / King County report ↗What we track and what remains unknown
Track employment among people experiencing homelessness, vehicle dwelling, rent relative to wages, evictions, and exits into stable housing separately. Compare like-for-like populations and methods over time. These studies document the gap between employment and housing security; they do not establish that working homelessness has risen nationally since the frozen August 20, 2026 baseline.
King County homelessness methodology ↗04 / HOUSEHOLDSSpending isn't the same as financial security.
What it measures
Aggregate consumer spending and saving show activity and resources across the economy.
What it misses
Totals can conceal differences between higher- and lower-income households, debt-funded consumption, or a loss of savings among households with fewer buffers. GDP, corporate earnings, and asset prices can improve even when gains are concentrated and median households have not recovered.
What we check alongside it
Inflation-adjusted consumption and earnings by income group, savings distribution, credit-card and auto delinquencies, household debt burdens, and median-versus-aggregate recovery.
BEA personal saving ↗05 / CREDIT ACCESSCredit stress can appear before missed payments.
What conventional measures miss
Delinquency rates count accounts already behind. They do not capture reduced credit limits, rejected applications, closed accounts, unaffordable refinancing, or households relying on relatives to stay current.
What we check alongside it
Credit-access surveys, lender tightening, account closures, credit-limit changes, application denials, and borrowing from family. Track whether access recovers alongside falling delinquency.
06 / CASCADING LOSSESOne financial shock can trigger several others.
What conventional measures miss
Bankruptcy, foreclosure, eviction, and vehicle repossession are reported separately. The same household may experience several of these losses in sequence; independent totals cannot show whether the losses are linked.
What we check alongside it
Bankruptcy filings by chapter, mortgage and rent distress, eviction filings, auto delinquencies and repossessions, and documented transitions between them. Do not assume causation or count the same episode as multiple independent shocks.
07 / COST OF STABILITYContinuing to function can consume the capacity to recover.
What conventional measures miss
Spending, production, or liquidity can hold steady while savings are exhausted, maintenance is deferred, borrowing grows, or emergency support becomes routine. Stable output does not show what sustaining it costs.
What we check alongside it
Changes in cash buffers, interest expense, debt refinancing, deferred capital spending, emergency interventions, and the time and resources required to restore ordinary operations.
How we interpret it
Stabilization means deterioration has slowed or stopped; absorption means activity continues despite pressure; repair requires evidence that capacity to withstand future pressure has been restored. These categories can coexist and must be supported by observations, not assigned automatically.
08 / SMALL BUSINESSESAn open business may still be financially fragile.
What conventional measures miss
Business survival and new registrations do not establish adequate owner pay, healthy margins, timely supplier payments, or the ability to operate without personal debt.
What we check alongside it
Small-business credit conditions, closures, payroll and staffing, tax arrears where available, owner borrowing, and post-restructuring viability.
09 / INFRASTRUCTUREService continuity can conceal declining redundancy.
What conventional measures miss
Electricity, water, transport, and other essential services can continue operating while maintenance backlogs grow and backup capacity shrinks. A single outage count cannot establish resilience.
What we check alongside it
Maintenance backlogs, reserve margins, restoration times, outage frequency and duration, capital replacement, and dependence on emergency measures.
10 / HEALTHCARE & MEDICAL DEBTHealthcare spending does not prove access to care.
What conventional measures miss
Medical costs may move into credit-card debt, loans, or unpaid household bills rather than medical collections. Falling out-of-pocket spending can reflect delayed treatment, skipped prescriptions, or unmet need rather than improved affordability.
What we check alongside it
Medical debt across credit products, delayed or forgone care, premiums and deductibles relative to income, insurance gaps, health-related work limitations, unpaid caregiving, hospital closures, uncompensated care, and provider operating margins.
Why this crosses domains
Lost access to care can affect employment, earnings, housing, household credit, and hospital finances. Track the sequence and distinguish documented effects from plausible pathways.
11 / MEDICAID ELIGIBILITYCoverage losses are not the same as reduced healthcare need.
What is changing
Under federal law, certain Medicaid expansion adults face work or community-engagement requirements starting January 1, 2027. Illinois generally describes an 80-hour qualifying-activity route or at least $580 in monthly earnings, with exemptions. Affected expansion adults also move to six-month eligibility renewals. Rules vary by eligibility category; many people, including those in other Medicaid groups, are not subject to these work requirements.
Why the counts need context
A decline in enrollment could reflect changed eligibility, a move to other coverage, failure to complete paperwork, or an erroneous termination. It does not by itself establish increased employment or reduced need. Illinois also identifies a shorter retroactive-coverage window beginning in 2027 and certain new cost-sharing requirements in 2028.
What we check alongside it
Enrollment by eligibility group, renewals, procedural terminations, reinstatements, exemptions, uninsured rates, forgone care, medical debt, uncompensated care, employment, and state administrative costs. Separate changes in law from measured outcomes after implementation.
Recent economic
evidence.
A selection of the newest observations from the past 14 days appears here. The complete, searchable record lives in the Evidence Ledger.
The full record includes every published observation, counterevidence, and correction.
Explore the complete Evidence LedgerAll published observations, including historical records
The baseline
came first.
Each new development helps us understand how pressure moves through the system, where it builds, what keeps things stable, and whether the patterns we're seeing can help us anticipate what happens next.
Freeze the claim before observing forward.
The prediction, baseline conditions, disruption threshold, and falsification criteria were fixed before subsequent observations entered this ledger.
Compare new evidence with the frozen baseline.
Later events are assessed against the prospective record rather than using hindsight to redefine what counts as support.
Preserve counterevidence.
Genuine normalization remains visible alongside evidence that challenges the prospective prediction.
Distinguish correction from systemic repair.
A local response can absorb or correct pressure without establishing that repair has been restored across the wider system.