A plain-English guide to reading ISM plus JOLTS as a single signal.

Today's Tuesday September 1 tape delivered the archetypal 10:00 AM ET first-of-the-month compressed release: ISM Manufacturing, JOLTS, and Construction Spending all inside the same one-minute window. The market treated the three prints as a single signal even though the surveys measure different things and the release windows are technically independent. That treatment has a mechanism, a set of tape-reading conventions, and a set of pitfalls that a first-Tuesday-of-the-month tape reader has to hold in mind. This piece is the framework for reading ISM plus JOLTS as a single signal: what the fusion is, when the two prints reinforce vs. contradict, and how the market prices the combined output on the fifteen minutes after the 10:00 AM release.

Why the market treats them as one signal

Three reasons the fusion is the reading convention rather than treating them independently.

1. Shared release window. Both prints land at 10:00 AM ET on the first business day of the month that also carries the ISM release (JOLTS often lands with a one-day lag; when the two land together the market treats them together). The one-minute simultaneity forces the tape to price the combined output in a single directional response rather than two sequential responses. This is a mechanical feature of the release calendar rather than an analytical judgment.
2. Complementary coverage of the labour market. ISM Manufacturing's Employment sub-index measures the demand-side signal for factory-sector labour (are manufacturers hiring); JOLTS measures the openings-and-flows side of the labour market economy-wide. The two are not redundant; they cover different slices of the same underlying question. When they agree, the signal on labour-market direction is stronger than either alone. When they disagree, the signal is that the labour market is in a mixed regime worth reading more carefully.
3. Both are Fed reaction-function inputs. ISM Manufacturing's Prices Paid sub-index is one of the specific inputs the Cleveland Fed and the New York Fed use in their inflation nowcasting frameworks; JOLTS's vacancy-to-unemployment ratio is one of the specific inputs to the Federal Reserve's assessment of labour-market tightness (Chair Powell cited it repeatedly during 2022-2024 and Chair Warsh's August 5 statement acknowledged it as an ongoing input). The two together, on the same morning, are unusually dense Fed-relevant data.

The four possible combined signatures

The two prints produce four archetypal combined signatures depending on the direction of each. Every first-Tuesday tape can be classified into one of the four.

1. Both firm (aligned hawkish). ISM Manufacturing above 53 with Employment above 50 and Prices Paid above 55; JOLTS openings above 7.0 million and quits stable. Today's Sep 1 print is this signature: 54.6 with Employment 51.2 and Prices Paid above 60; JOLTS 7.271M with quits 3.1M unchanged. Tape response: dollar bid, front-end yields up 6-10bp, gold and safe-havens offered, curve either flat or steepening slightly. Response magnitude typically 1.5x the sum of the two prints' individual expected magnitudes; that is the compression signature.
2. Both soft (aligned dovish). ISM below 50 with Employment below 48 and Prices Paid falling; JOLTS openings below 6.8 million with a hires collapse. Tape response: dollar offered, front-end yields down 8-12bp, gold and safe-havens bid, curve typically bull-steepens. Magnitude 1.5x-2x sum. Under this signature the market treats the labour-and-manufacturing complex as decisively softening; Fed easing pricing shifts materially inside the first 30 minutes.
3. ISM firm, JOLTS soft. The mixed signature that most confuses the tape. Manufacturing sector expanding, wider labour market cooling. Response is muted in aggregate but internally two-way: the growth read is neutral-firm, the labour read is neutral-soft, and the Fed reads them differently. Typical response: dollar mixed, front-end yields little changed, gold flat, curve steepens as the labour softness dominates the belly.
4. ISM soft, JOLTS firm. The other mixed signature. Manufacturing contracting, labour market still tight. Typically the tape reads this as stagflationary: gold bid on the persistent-labour-tightness inflation risk, front-end yields firm (Fed cannot ease while labour is tight), long-end mixed. Rare but historically the most durable in follow-through when it happens.

The five diagnostic dimensions

Given a first-Tuesday print, five dimensions to check for combined-signal attribution.

1. ISM Employment sub-index versus JOLTS hires rate. These two measure similar things from different angles. Aligned in direction (both firming or both softening) is a clean signal on labour-market direction. Opposed (Employment sub-index firming while JOLTS hires rate falling) is the frozen-market signature and requires the Beveridge-curve reading to interpret. Today's print: ISM Employment softened (51.2 from 52.8), JOLTS hires rate softened (3.2 percent from 3.4). Aligned soft on the labour side, even as the aggregate reads firm through openings and PMI.
2. ISM Prices Paid versus JOLTS quits rate. Prices Paid is the inflation-pressure input; quits rate is the wage-pressure input (workers quit for higher pay when the labour market is tight). Aligned firm on both is the specific stagflation-risk signature. Today's print: Prices Paid above 60 (firm), quits 3.1M stable (neutral). Not the stagflation signature; a firm-prices, stable-wages combination.
3. ISM New Orders trajectory versus JOLTS openings trajectory. New Orders is the forward-looking manufacturing demand signal; openings is the forward-looking labour demand signal. Aligned direction on both is the cleanest forward-looking growth signal. Today's print: New Orders down 3.0 to 53.7 (still expansion, but softer trajectory); openings up 89k (firmer trajectory). Slightly mixed; the manufacturing forward-look is softening, the labour forward-look is not.
4. ISM overall trajectory versus JOLTS openings-to-unemployment ratio. The macro-scale reads on manufacturing and labour respectively. On today's print, PMI at 54.6 (expansion) and vacancy-to-unemployment ratio still above 1.0 (tight); both fit the hawkish-Fed reaction function's growth-supportive frame.
5. Relative surprise magnitude. Which print surprised more relative to consensus. On today's print, ISM was close to consensus (54.6 vs 54.8), JOLTS was above consensus (7.271M vs 7.0M expected). The surprise was labour-side; the market response should have weighted labour more than manufacturing. The tape did exactly that, with the front-end 2-year moving 7bp on a compressed release where the manufacturing read was in line.

The Warsh-era reading convention

Under Warsh's hawkish reaction function (committed at Jackson Hole on August 28), the aligned-firm signature (Case 1) reads as reinforcing the September FOMC hike case; the aligned-soft signature (Case 2) is the specific input that would push back on the hike case; the two mixed cases (Cases 3 and 4) leave the September FOMC pricing where it is. Today's print at Case 1 with the labour-side softness caveat gets read as roughly 80 percent of a full hawkish-corroboration signal; the market priced roughly that (2-year up 7bp on a session where a full corroboration would have moved it 10-12bp). The distribution-flip framework called for the Week 2 data-flow test to be corroborating or contradicting; today was corroborating with a modest caveat.

The three most common misreads

1. Reading the Construction Spending print as a third signal. Construction Spending lands in the same window as ISM and JOLTS but is backward-looking (July data on the September 1 release, versus the August data ISM covers and the July data JOLTS covers). The market treats it as a footnote; a tape reader who elevates it to a third signal will over-fit to a print the desk is discounting.
2. Reading the ISM Prices Paid as a demand signal. Prices Paid is an input-price index (what manufacturers are paying for materials, wages, energy). It is not a consumer-price signal or a demand signal. A Prices Paid print above 60 is an inflation-pressure input to the Fed's reaction function, not evidence of firm demand. On today's print the market read Prices Paid correctly; a common desk-level misread is to treat it as a growth signal.
3. Ignoring the JOLTS revisions. The JOLTS release includes revisions to the prior two months. Large revisions can change the trajectory of the series and change the market's read of the current print. Today's print revised June from 7.437M to 7.182M (a 255k downward revision), which changed the openings-trajectory read from firming to roughly flat. The desk that reads the current print without the revisions will over-read the strength of the labour market.

Related reading

• Today's analysis piece: the Tuesday tape that motivates this framework.
• Compressed macro release days: parent framework on multi-print sessions.
• The Beveridge curve: the framework for reading the frozen-market signature the JOLTS hires-softness produced.
• A Chair keynote flipping the distribution: the Friday framework that today's print corroborated.