What a Tighter Standard Actually Changes
In the U.S. credit-card market, a tighter standard is not a switch that directly subtracts a fixed number of cards from the economy. It is a change in a lender's policy for deciding which applications to approve, which prospective borrowers to approach, and what terms to offer. The useful judgment is therefore to follow the policy change through those decisions before interpreting what happens to inquiries, failed applications, or new accounts.
A survey response is a direction, not a card-count forecast
The Federal Reserve's Senior Loan Officer Opinion Survey on Bank Lending Practices, usually called SLOOS, asks banks how their credit-card approval standards changed over the previous three months. It asks a separate question about demand, apart from normal seasonal variation. That separation matters because a bank can become more selective even when consumers still want credit, or consumers can lose interest even while a bank is willing to approve more applications.
For standards, the survey uses five response categories: eased considerably, eased somewhat, about unchanged, tightened somewhat, and tightened considerably. The study turns those categories into a simpler directional variable: -1 for easing, 0 for no change, and 1 for tightening. This coding is useful for comparing many banks across time. It is not a percentage scale. A response of 1 does not mean that approval odds fell by one percent, nor does it imply that two banks reporting “tightened” made changes of equal size.
That difference between direction and magnitude is easy to miss when a chart shows bars rising above or below a zero line. The bars summarize how many respondents reported moving in each direction; they do not measure a common physical unit of restrictiveness. The Federal Reserve paper cautions that the interpretation of qualitative responses can vary across banks and over time. It also notes that banks may be reluctant to report easing, which could make the observed contrast between tightening and easing different from the underlying contrast in policy.
SLOOS is valuable precisely because it asks lenders about a decision that an account count can only reveal indirectly. The survey has covered major loan categories since the early 1990s, and the credit-card sample generally includes nearly 60 U.S. commercial banks. Those banks span the Federal Reserve's districts and hold about 70 percent of assets at domestically chartered banking institutions, according to the paper. But the sample is weighted toward large banks, participation is voluntary, and the study uses a particular historical panel. The survey is broad enough to be informative about bank lending conditions; it is not a complete census of every issuer or a direct measure of every consumer's approval probability.
The survey's time window is part of its meaning. A response describes how the bank says its policy changed during the preceding three months, not the full history of its underwriting rules and not necessarily the policy faced by every applicant in every product. The study's directional coding makes it possible to compare a bank-quarter response with an observed change in that bank's offers, but it deliberately discards the difference between “tightened somewhat” and “tightened considerably.” That trade-off buys comparability while limiting precision. A reader can reasonably treat a positive tightening indicator as evidence that the bank reported movement in a restrictive direction; they should not treat it as an instruction to calculate a fixed number of lost accounts.
Demand is the crucial companion to the standards response. Imagine that a bank reports tightening while consumer demand strengthens. Applications may still rise because more people are trying to obtain cards, even if the bank rejects a larger share or approaches fewer people. Reverse the conditions—unchanged standards and sharply weaker demand—and new accounts may fall without any restrictive policy move. A standards response identifies a lender-side change; it does not settle the downstream outcome by itself.
The same caution applies to the word “tightening.” In ordinary language, it may suggest one broad action, but a bank can implement stricter access in several ways. It might reduce the number of solicitations, stop approaching some prospective borrowers, raise an internal screening threshold, or retain the same broad customer strategy while offering smaller limits or less favorable prices. The survey's general standards question records the direction of the underwriting policy. It does not specify which operational lever moved.
A useful thought experiment is to hold consumer demand fixed and change only the lender's policy. If the bank mails fewer offers, the first visible effect is a reduction in opportunities to respond. If it keeps the same mailing volume but screens more strictly, inquiries may continue while fewer applications transition into new credit. If it keeps both outreach and approval rates broadly stable but reduces limits, the number of new cards may barely change even though the amount of credit made available per account does. The word “tightening” covers all three possibilities at the survey level, while the market consequences differ. That is why an action measure and a transition measure are needed to interpret the survey response.
From a bank's decision to a consumer-facing measure
The Federal Reserve study uses mailed offers to observe one of those operational choices. Mintel Comperemedia collects credit-card solicitations forwarded by participating consumers. The records contain more than the existence of a mailing: they can include the minimum credit limit and the regular, non-promotional purchase interest rate—the “go-to” rate. A bank controls whether to mail an offer, how much mail to send, and what terms the offer displays. That makes mail volume and offer terms action-based signals of supply.
The phrase “action-based” does not mean “complete.” Banks acquire credit-card customers through several channels, including the internet, direct mail, telemarketing, and direct selling. The study's evidence is especially useful because personal credit cards rely heavily on direct mail and because the observed mailing is a decision made by the lender before the consumer opens an account. It does not establish that a bank's unobserved online or direct-selling activity moved in exactly the same way. A decline in mailed offers is evidence about the mail channel; it is not, standing alone, a count of all offers available to consumers.
It is also important to distinguish a volume measure from a stock measure. The number of offers mailed during a quarter describes activity in that period; it does not tell us how many offers remain available, how many recipients accept them, or how much credit is already outstanding. Offer terms add another dimension. Two banks could mail the same number of solicitations while differing substantially in minimum limits or regular purchase rates. Conversely, a bank could mail fewer offers but make each remaining offer more generous. The value of the Mintel measure is that it lets the researcher observe these decisions separately instead of forcing every form of supply into one count.
The data also reflect a sampling process. Mintel selects about 4,000 consumers each month from a larger pool, and about 2,500 on average participate. Participants forward the material from tracked commercial mail, allowing the researchers to reconstruct offer volume and terms. Weights based on demographic and socioeconomic information help estimate national mail volume. Credit-score information was merged for offer recipients beginning in 2007, which makes it possible to examine whether the relationship differs across prime, near-prime, and subprime groups in the portion of the sample where those scores are available. These details do not invalidate the measure; they identify what the measure is actually measuring.
The downstream CFPB indicators ask different questions again. The inquiry index counts the monthly number of consumers with credit-card inquiries and expresses it relative to January 2010. An inquiry is a hard credit pull or comparable screening event associated with an attempt to obtain credit. It is not an approval and not an account opening. The CFPB uses an index because not all lenders pull credit reports from all three nationwide consumer reporting agencies. An indexed series can show that activity is higher or lower than its base-period level without claiming to observe every raw inquiry in the market.
The base makes the index relative rather than mysterious. If the seasonally adjusted inquiry index were 120, that would mean the measured level was 20 percent above its January 2010 reference level; it would not mean that 120 percent of consumers applied, or that inquiries had increased by 120 percent. The index also does not give the reader a direct denominator for the whole population. Its strength is comparison across months within the same construction, provided the reader remembers the agency-coverage and data-finality limits.
Seasonality adds another layer. Consumers may apply for credit more often at some times of year than others, independent of a durable change in underwriting. The CFPB therefore displays a seasonally adjusted series alongside an unadjusted series. The adjusted series is useful when the question concerns an underlying movement rather than a predictable calendar pattern. The unadjusted series remains useful when the question is what actually occurred in each month. Neither version removes the need to identify the time window and the reporting coverage.
The credit-tightness index sits after the inquiry. It measures consumers who applied for a credit card and did not obtain additional credit, also relative to January 2010. An increase means that fewer inquiries are transitioning into loans compared with the base period. The CFPB adjusts the data to hold applicants' credit scores constant. The purpose is to keep a shift in the risk mix of applicants from being mistaken for a shift in lender policy. If a wave of applications contains more high-risk borrowers, a lower transition rate might otherwise look like a sudden tightening even if lenders did not change their rules.
This makes credit tightness a conditional comparison, not a simple tally of refusals. The reader is asking how the transition from inquiry to additional credit changes after accounting for one important feature of the applicant pool. That is more informative than comparing two raw months whose applicants differ sharply in credit scores, but it still does not mean the remaining difference has one identifiable cause. The measure is designed to make lender-side interpretation more plausible, not to turn an observational index into a direct survey of underwriting rules.
That adjustment is informative but not magical. Holding credit scores constant does not hold every applicant characteristic constant, and the index does not identify a unique policy lever. The index tells us about a transition outcome after an inquiry; it does not reproduce the wording of a bank's SLOOS response.
The relationship among the measures is best understood as a sequence of decisions with possible drop-off at each step. A lender's reported standard is a qualitative signal about policy. A mailed offer records one controlled form of outreach. An inquiry records an attempt or screening event. The tightness measure records that an attempt did not become additional credit, after a credit-score adjustment. An origination records the completed account outcome. A movement can be visible at one point and muted at another because a later decision, a different channel, or a different population offsets it.
There is a second source of confusion: these measures use different units. A standards response is a category; mail volume is an observed activity measure; an inquiry index is relative to a base month; credit tightness is a relative transition measure; and originations are counts or credit-limit amounts. A percentage change in one cannot be compared with an index-point change in another as though they shared a denominator. Even when two indicators are both described as “up,” they may be measuring different events, for different people, in different months.
This sequence is a conceptual map, not a claim that every bank follows one identical timetable. SLOOS asks about changes over a prior three-month period, while the public indicators are assembled from monthly consumer records. Still, the order clarifies why the measures cannot be treated as interchangeable. A reduction in solicitation can affect opportunities before a hard-pull measure records an attempt; an inquiry may not result in additional credit; and an account can be opened with a smaller limit or a higher rate than another borrower sees. “Access” has several observable dimensions.
That is also why a tighter standard can matter before the origination count changes. If the lender first reduces mail volume or narrows the set of recipients, the effect may appear as fewer opportunities rather than an immediate fall in all new accounts. If other banks expand their offers, or if consumers respond more strongly to the remaining offers, aggregate originations may soften less than the restricted lender's supply. Conversely, a higher failure rate among inquiries can coexist with strong origination growth if the volume of applications or the number of participating lenders increases enough.
A careful reading therefore keeps two kinds of statements separate. “Banks reported tighter standards” describes what survey respondents said. “Mail volume was lower” describes an observed lender action in a particular channel. “Credit tightness increased” describes a population-level transition outcome after an inquiry and a score adjustment. “Originations rose” describes completed new accounts. Connecting those statements is an empirical task, not a matter of substituting one noun for another.
The next test is correspondingly specific: when the same banks that report tighter standards are observed in the mail market, do their actions show less solicitation? The answer can be measured, but only if the comparison preserves the lender, the period, and the distinction between supply and demand.