ISM vs. S&P Global PMI: What a Reading Above 50 Means

PMI surveys ask business respondents whether selected conditions changed from the previous month. Depending on the survey, questions cover items such as orders, output or business activity, employment, supplier delivery times, inventories, and prices. The result is a compact description of the direction of change across respondents, not a census of all production, sales, or jobs. ISM describes its manufacturing and services reports as data compiled from purchasing and supply executives; S&P Global describes its surveys as monthly business-condition measures. Their [ISM reports overview]({source:ismPmiReportsOverview}) and [PMI FAQ]({source:spGlobalPmiFaq}) explain their respective scopes.

The word “purchasing” can be misleading when read literally. Purchasing managers may report useful information about orders and suppliers, but the survey also asks about broader operating conditions. A PMI value does not say that purchasing volume, factory output, or the overall economy grew by that many percent. It summarizes the share and direction of respondents' answers after the provider applies its own index method.

The term PMI is also used for several different series. ISM publishes a Manufacturing PMI and a Services PMI. S&P Global publishes manufacturing and services surveys, and its services headline is the Business Activity Index. A report headline may therefore be a composite of several components or one particular component. Check the provider, sector, and exact index name before comparing two numbers.

For a typical higher-or-same-or-lower question, a diffusion index is calculated as the percentage reporting an increase plus one-half of the percentage reporting no change. In symbols: index = % higher + 0.5 × % unchanged. Respondents reporting a decrease contribute zero to that particular calculation. If equal shares report higher and lower, the index is 50 regardless of how many report no change.

Suppose 32% of respondents report more new orders, 46% report no change, and 22% report fewer orders. The index is 32 + 0.5 × 46 = 55. It is above 50 because the share reporting increases exceeds the share reporting decreases. If the responses were 22% higher, 46% unchanged, and 32% lower, the index would be 45. The calculation uses the direction of responses, not the size of each company's change.

Above 50 is commonly described as expansion or an increase in the activity named by that component; below 50 is described as contraction or a decrease. The interpretation belongs to the question. For example, an employment component above 50 indicates a net balance of respondents reporting higher employment, while a prices component above 50 indicates more respondents reporting rising prices. It does not mean every firm expanded or that the overall economy grew by the same percentage.

A diffusion index records how widely a change is reported, not how large the change was. One respondent who reports a small increase and one that reports a very large increase enter the response categories as increases; the PMI does not add their output changes together to estimate a national growth rate. A reading of 55 therefore means the response balance is five index points above its neutral line. It does not mean output rose 5% or that 55% of companies expanded.

The index also does not reveal every reason for a response. Slower supplier deliveries can occur when demand is strong and suppliers are busy, but delays can also reflect capacity limits, weather, transport problems, or a temporary disruption. A component summarizes what respondents report; it does not by itself distinguish a demand boom from a supply bottleneck. Read the component's own wording and any response details in the release.

ISM labels its index readings with percent signs in its releases, but the figure remains an index rather than a percentage growth rate. The [ISM methodology pages]({source:ismPmiManufacturingMethodology}) describe the diffusion calculation and how the manufacturing headline is formed. Use the 50 threshold to describe the direction of respondents' answers, then use official production, employment, or national-account data to measure realized output and activity. BEA's [GDP series]({source:beaGrossDomesticProductOverview}) measures a different object from a business survey index.

The ISM Manufacturing PMI combines five component indexes with equal weights: New Orders, Production, Employment, Supplier Deliveries, and Inventories. Each contributes one-fifth to the composite before the seasonal-adjustment procedure described by ISM. The [Manufacturing PMI report]({source:ismPmiManufacturingMethodology}) describes the five components and the 50-point interpretation. It also provides other subindexes—such as prices, imports, and backlog—that do not all enter the headline composite.

Supplier Deliveries needs special care. A value above 50 means deliveries are slower, not faster. In the ISM composite, slower deliveries can accompany stronger demand and busier suppliers, so this component is oriented differently from a simple “faster is better” service measure. But slow deliveries can also reflect a supply constraint. The headline blends this component with orders, production, employment, and inventories; it cannot tell from the composite alone why delivery times changed.

Seasonal adjustment matters too. ISM adjusts the Manufacturing PMI headline and selected component series; its public method identifies New Orders, Production, Employment, and Inventories for adjustment, while Supplier Deliveries is handled separately. ISM publishes seasonal factors and can update them as its annual process requires. Read the footnote to the monthly release to see which series are seasonally adjusted and whether the comparison uses the same data vintage. The [seasonal-adjustment page]({source:ismPmiSeasonalAdjustmentFactors}) posts the current factor information.

S&P Global's manufacturing headline combines five diffusion indexes too, but it does not use ISM's equal 20% weights. S&P Global's published weights are 30% New Orders, 25% Output, 20% Employment, 15% Supplier Delivery Times, and 10% Stocks of Purchases; the supplier-delivery component is oriented for use in the composite. Its [PMI FAQ]({source:spGlobalPmiFaq}) explains those weights and the survey questions.

The surveys also use different respondent panels and weighting practices. ISM compiles data from its purchasing and supply executive panels, organized around industries' contribution to GDP. S&P Global describes its panels as structured by sector and company size. The [provider comparison]({source:spGlobalIsmPmiComparison}) sets out methodological differences, including respondents, panel composition, seasonal adjustment, and headline calculations. Since this comparison is published by S&P Global, use it alongside ISM's own report notes when checking ISM-specific details.

These differences help explain why two manufacturing readings for the same month can diverge. New orders may gain weight in one composite, while an equal-weight index gives production, inventories, and delivery times the same arithmetic influence. A change in which firms answer, their industry mix, seasonal adjustment, or survey cutoff can also affect the release. Both indexes can still be useful as separate indicators; their difference is not a direct estimate of how much one survey overstates activity.

The headline definitions differ even more for services. ISM's Services PMI is a composite built from Business Activity, New Orders, Employment, and Supplier Deliveries. Its [August 2026 Services report]({source:ismPmiServicesMethodology}) identifies these components and explains that an above-50 Supplier Deliveries reading means slower deliveries. ISM's survey also spans a broad set of nonmanufacturing industries, including some industries that another provider may classify or exclude differently.

S&P Global's headline for its services survey is the Business Activity Index, not a four-component composite called a Services PMI. That index summarizes reported changes in business activity among its services respondents. S&P Global's [comparison with ISM]({source:spGlobalIsmPmiComparison}) notes differences in sector coverage and survey design. As a result, comparing ISM Services PMI with S&P Global Business Activity is not a comparison of two identically defined composites.

If you want to compare service-sector survey signals, first decide whether you mean activity, new business, employment, or a broader composite. Then confirm whether the survey covers only private services or a wider set of nonmanufacturing industries, whether its headline is a composite, and how supplier deliveries enter. The chart label “services PMI” can conceal these differences, so retain the formal series name in notes and analysis.

<!-- learn:illustration -->

Assume, only to isolate the arithmetic, that both manufacturing surveys reported these same component values for one month: New Orders 55, Output or Production 52, Employment 48, Supplier Deliveries 58, and Inventories 47. These are invented inputs, not current observations, and real providers do not necessarily have matching questions or respondent panels.

An equally weighted composite would be (55 + 52 + 48 + 58 + 47) ÷ 5 = 52.0. Applying S&P Global's stated weights to those same hypothetical values gives (0.30 × 55) + (0.25 × 52) + (0.20 × 48) + (0.15 × 58) + (0.10 × 47) = 52.5. The half-point difference comes from weights alone in this simplified calculation. The result does not claim that the actual ISM and S&P Global releases used identical component values.

The example also shows why the headline does not erase the component story. New Orders is above 50 and Employment is below 50; an overall reading just above 50 does not mean every part of manufacturing is expanding. A useful reading reports both the composite and the component pattern, then explains the different denominators and survey methods before comparing providers.

Start by selecting the question. Use manufacturing components to examine surveyed goods-producing businesses, and use the relevant services activity or composite index for service-sector conditions. For broader U.S. economic output, consult measures such as BEA GDP; for jobs, use labor-market releases. A PMI can offer timely information about the direction and breadth of surveyed changes, but it is not a substitute for official data that measure output, employment, prices, or spending under their own definitions.

For month-to-month analysis, compare a series with its own prior readings and retain its release month, exact index name, seasonally adjusted status, and vintage. Avoid subtracting an ISM value from an S&P Global value as if their difference were a shared unit. If their directions conflict, inspect question wording, sector coverage, weights, response timing, and subindexes. A component may also be near 50, where small response changes can shift its classification without a large change in real activity.

Neither an above-50 reading nor a run of readings guarantees GDP growth, a hiring surge, or a recession. Use PMIs as one layer of evidence, then check whether measured production, income, employment, and spending confirm the interpretation. For a broader discussion of measured output, see nominal vs. real GDP. For household demand, see retail sales vs. PCE spending. Those series answer different questions from managers' month-to-month survey responses.