New Orders, Shipments and Production Measure Different Things

Question

What can the Census Bureau M3 survey and the Federal Reserve G.17 release each establish about national manufacturing activity, and what can neither establish?

Who can use this

• Analysts, journalists, researchers and public-sector staff who need to read national manufacturing statistics without treating orders, shipments and production as interchangeable.
• Authors checking whether a sentence about factory activity is supported by the statistic cited for it.

Chronology

1. 2026-09-02 Census Bureau publishes the full M3 report for July 2026.
2. 2026-10-02 Census Bureau publishes the full M3 report for August 2026, reporting new orders up $0.7 billion to $663.5 billion, which implies a July level of about $662.8 billion against the $663.6 billion first published.
3. 2026-10-16 Federal Reserve publishes the G.17 release with September 2026 estimates.
4. 2026-11-03 Full M3 report for September 2026 is scheduled.
5. 2026-11-24 Federal Reserve plans its annual G.17 revision, moving the index base year to 2022.

Claims

Observed

C-01 In the M3 survey a new order is a stated intention to buy, and the delivery may be immediate or later. U.S. Census Bureau
Source text: “A new order is a communication of an intention to buy for immediate or future delivery.”

C-02 M3 shipments measure the dollar value of products sold, not a count of goods made. U.S. Census Bureau
Source text: “shipments measure the dollar value of products sold by manufacturing establishments”

C-03 The published universe estimate of new orders is derived from shipments and the change in unfilled orders. U.S. Census Bureau
Source text: “The universe estimate of new orders is derived from the monthly estimate of shipments plus the change in unfilled orders between the current and prior period.”

C-04 The Census Bureau gives incomplete reporting of new orders as the reason the standard ratio procedure is not used for that item. U.S. Census Bureau
Source text: “The standard ratio estimate procedure is not followed for new orders because not all companies report new orders”

C-05 Orders data are not requested from industries where deliveries are mostly immediate. U.S. Census Bureau
Source text: “Orders data are not collected from industries that have mostly immediate deliveries.”

C-06 The M3 panel is not a probability sample, so its sampling error cannot be measured. U.S. Census Bureau, U.S. Census Bureau
Source text: “The M3 panel is not based on a probability sample; therefore, the sampling errors that are normally provided with sample surveys cannot be measured.”

C-07 M3 headline figures are seasonally adjusted but are not adjusted for inflation. U.S. Census Bureau
Source text: “Figures in text are adjusted for seasonality, but not for inflation.”

C-08 New and unfilled order figures leave out semiconductor manufacturing. U.S. Census Bureau
Source text: “Figures on new and unfilled orders exclude data for semiconductor manufacturing.”

C-09 As first published, July 2026 new orders for manufactured goods were $663.6 billion. U.S. Census Bureau
Source text: “New orders for manufactured goods in July, up following two consecutive monthly decreases, increased $5.8 billion or 0.9 percent to $663.6 billion”

C-10 As first published, July 2026 shipments were $658.8 billion. U.S. Census Bureau
Source text: “Shipments, up nine of the last ten months, increased $5.3 billion or 0.8 percent to $658.8 billion.”

C-11 As first published, unfilled orders rose $9.9 billion in July 2026 to $1,600.3 billion. U.S. Census Bureau
Source text: “Unfilled orders, up twenty-four of the last twenty-five months, increased $9.9 billion or 0.6 percent to $1,600.3 billion.”

C-12 The G.17 production index measures real output relative to a base year. Board of Governors of the Federal Reserve System
Source text: “The production index measures real output and is expressed as a percentage of real output in a base year”

C-13 Federal Reserve capacity is a concept of sustainable maximum output under a realistic work schedule. Board of Governors of the Federal Reserve System
Source text: “The greatest level of output an industry can maintain within the framework of a realistic work schedule, after factoring in normal downtime”

C-14 The G.17 capacity utilization rate is a ratio of the output index to the capacity index. Board of Governors of the Federal Reserve System
Source text: “The rate of capacity utilization equals the seasonally adjusted output index expressed as a percentage of the related capacity index.”

C-15 Where physical product data are not available, G.17 output is estimated from production-worker hours. Board of Governors of the Federal Reserve System
Source text: “When suitable data on physical product are not available, estimates of output are based on production-worker hours by industry.”

C-16 The first G.17 estimate for a month rests on about 78 percent of the source data, by value added. Board of Governors of the Federal Reserve System
Source text: “about 78 percent of the source data (in value-added terms) are available”

C-17 The Federal Reserve industrial sector covers manufacturing, mining and electric and gas utilities. Board of Governors of the Federal Reserve System
Source text: “defined by the Federal Reserve to comprise manufacturing, mining, and electric and gas utilities”

C-18 G.17 manufacturing also includes logging and certain publishing industries that NAICS classifies elsewhere. Board of Governors of the Federal Reserve System
Source text: “Manufacturing comprises NAICS manufacturing industries (sector 31-33) plus the logging industry and the newspaper, periodical, book, and directory publishing industries”

C-19 In the October 16, 2026 release, manufacturing capacity utilization for September was 75.9 percent. Board of Governors of the Federal Reserve System
Source text: “Capacity utilization for manufacturing increased 0.1 percentage point in September to 75.9 percent”

C-20 The Federal Reserve Board benchmarks its monthly capacity estimates to the Census Bureau QPC. U.S. Census Bureau
Source text: “uses QPC data to benchmark monthly estimates of capacity output and utilization”

Derived

C-21 Published M3 new orders are not an independent reading alongside shipments, because the estimate is built from shipments and the change in unfilled orders. U.S. Census Bureau
Basis: Follows directly from C-03; a revision to shipments or unfilled orders moves the new orders estimate.

C-22 July 2026 shipments plus the rise in unfilled orders come to $668.7 billion, which is $5.1 billion above the $663.6 billion of new orders first published. U.S. Census Bureau
Basis: Arithmetic on C-09, C-10 and C-11: 658.8 + 9.9 = 668.7, and 668.7 - 663.6 = 5.1, using rounded published figures.

Inferred

C-23 The $5.1 billion difference is consistent with shipments including the semiconductor industry while new and unfilled orders exclude it, but the release does not state this reconciliation. U.S. Census Bureau
Basis: The report states in C-08 and its table notes that semiconductor data are in shipments aggregates and out of orders aggregates. No Census text read for this brief quantifies the gap, so the link remains an inference.

C-24 An M3 change and a G.17 change in the same month are not comparable on price, because one is nominal dollars and the other is real output. U.S. Census Bureau, U.S. Census Bureau, Board of Governors of the Federal Reserve System
Basis: C-02 and C-07 describe M3 values not adjusted for inflation; C-12 describes the production index as real output. The two can therefore differ with no change in underlying activity.

C-25 Neither a rise in new orders nor a rise in unfilled orders establishes when goods will ship or be produced. U.S. Census Bureau, U.S. Census Bureau
Basis: C-01 allows immediate or future delivery with no date attached, and the methodology states the estimate includes the effects of cancellations and modifications of previously reported orders.

C-26 A capacity utilization rate does not by itself separate a change in output from a change in estimated capacity, and it says nothing direct about orders. Board of Governors of the Federal Reserve System, Board of Governors of the Federal Reserve System, U.S. Census Bureau
Basis: C-14 defines the rate as a ratio of two estimated indexes, and C-13 and C-20 show capacity is itself a modeled and benchmarked estimate.