Industrial Distribution of Representation Elections Held in Cases Closed, Fiscal Year 2002

p. 138. The crop ends immediately after the "31-Manufacturing" subtotal row (double rule at the bottom); the table itself continues onto the following page, so later industry groups and the grand total are not on this image. Footnote markers 1 (on the stub heading "Industrial Group") and 2 (on "Total elections") are printed, but no footnote text appears on this crop, so `footnotes` is empty. Digit grouping is inconsistent in the source: most large numbers are printed without commas (e.g. 10366, 12711, 1322, 1843), but the "Total valid votes cast" cell of the "31-Manufacturing" row is printed with a comma as 11,329. Both are transcribed exactly as printed. The indented rows "Agriculture, Forestry, Fishing, and Hunting", "Mining", "Construction" and "31-Manufacturing" are sector subtotal rows; "Utilities" is a one-line sector set off by rules above and below. All subtotals were checked and agree with the sum of their component rows, and within every row the won/no-representative and union/no-union splits add to their totals, so no misprints were detected. Two row labels wrap onto two printed lines and are joined here: "Support Activities for Agriculture and Forestry", "Agriculture, Forestry, Fishing, and Hunting", "Building, Developing and General Contracting", and "Beverage and Tobacco Product Manufacturing". The last column heading is broken as "Eligible employ-ees in units choos-ing rep-resentati on" and is joined to "Eligible employees in units choosing representation".

p. 139. This crop is a continuation page ("—Continued"): it begins mid-table with no preceding sections and the footnote texts for the markers 1 (on "Industrial Group") and 2 (on "Total elections") are not printed on this image, so `footnotes` is empty. Suspected misprint in the source: for "Petroleum and Coal Products Manufacturing" the "Number of employees eligible to vote" is 964 while "Total valid votes cast" is 1199 — more votes cast than employees eligible. Both are transcribed exactly as printed; no correction was made. All three subtotal rows (32-Manufacturing, 33-Manufacturing, Wholesale Trade) were checked against the sums of their component rows in all 14 numeric columns and agree exactly, which supports the column alignment of the detail rows. Rules in the image separate the three sections (double rules above/below the 32-Manufacturing and 33-Manufacturing subtotal lines); the subtotal labels are indented in the stub but carry the wording shown. Leader dots after row labels were dropped; "Fabricated Metal Product Manufacturing" and "Transportation Equipment Manufacturing" are printed without leader dots. Multi-line stub labels (e.g. "Petroleum and Coal Products Manufacturing", "Electrical Equipment, Appliance and Component Manufacturing") were joined into one line. The heading "Eligible employees in units choosing representation" wraps with hyphenation across six printed lines ("employ- ees in units choos- ing rep- resentati on") and was rejoined; likewise "AFL-CIO unions", "Total elections", and "Number of elections in which no representative was chosen". No cells were illegible.

p. 140. The crop is a continuation page ("—Continued"); the column headings are reprinted here, so header_rows is filled. Footnote markers 1 (on "Industrial Group") and 2 (on "Total elections") appear in the stub/column headings, but no footnote text is printed on this crop, so `footnotes` is empty. Two subtotal rows are both labeled "Retail Trade" with different NAICS prefixes (44 and 45); transcribed as printed. Possible misprint in the source, transcribed as printed: for "Nonstore Retailers" the "Total valid votes cast" (153) exceeds the "Number of employees eligible to vote" (147). The 45-Retail Trade subtotal is consistent with 153 (108+1185+403+153 = 1849), so the digits are read correctly as printed. All other row and column cross-checks (union subtotals = AFL-CIO + other national + other local; union votes + no-union votes = total valid votes cast; the 44-, 45- and 48- subtotals against their component rows) add up exactly, which supports the digit readings. Only three digit-grouping commas are printed in the body, both in the 48-Transportation row (14,151 and 11,583); larger four-digit figures elsewhere in the table are printed without commas, as transcribed.

p. 141. Continuation page of Table 16; no footnote texts are printed on this crop, although the headings carry markers (Industrial Group^1, Total elections^2) whose definitions appear elsewhere in the table. The table continues past the bottom edge of the crop (the last printed row here is "Management of Companies and Enterprises"). Indented rows ("49-Transportation", "Information", "Finance and Insurance", "Real Estate and Rental and Leasing") are the printed industry-sector subtotal lines; they are transcribed as ordinary rows in place. The last column's heading is set with end-of-line breaks as "Eligible employ-ees in units choos-ing rep-resentati on" and is joined here as "Eligible employees in units choosing representation". Two printed oddities, transcribed as printed and not corrected: for "Funds, Trusts and Other Financial Vehicles (U.S. Only)" the total valid votes cast (212) exceeds the number of employees eligible to vote (183), and the same holds for the "Finance and Insurance" subtotal (319 vs 304). Also as printed, "Motion Picture and Sound Recording Industries" shows 0 elections won by unions yet 10 eligible employees in units choosing representation. All digits were legible; no doubtful cells.

p. 142. This crop is a continuation page ("—Continued"); the column headings are reprinted on it, so they are given here. Footnote markers appear in two headings ("Industrial Group^1" and "Total elections^2"), but no footnote text is printed on this crop — it appears on an earlier or later page of the table. The table also continues past the bottom rule of this crop (no grand-total row is shown here). Row structure: rows are grouped by NAICS sector, with an indented sector subtotal line closing each group ("Administrative and Support, Waste Management and Remediation Services", "Health Care and Social Assistance", "Arts, Entertainment and Recreation", "Accommodation and Foodservices", "Other Services (except Public Administration)"). "Educational Services" is printed with the subtotal indentation and sits alone between two double rules; it is transcribed with its printed label. Heading reconstruction: the last column is printed as "Eligible employ-ees in units choos-ing rep-resentati on", with the final word broken awkwardly across lines; it is joined as "Eligible employees in units choosing representation". Verification: every row's internal arithmetic is consistent as transcribed — the union-win components (AFL-CIO + other national + other local) sum to the "Total" won; the union-vote components sum to the union-vote total; union votes + votes for no union equal total valid votes cast; and each sector subtotal equals the sum of its member rows in all 14 numeric columns. No printed value appears to be a misprint. All numbers in the body are plain integers with no footnote markers attached; digit-grouping commas appear only in five-digit-and-larger values, as printed.

p. 143. This is the final (---Continued) page of Table 16; the earlier industrial groups appear on the preceding page(s) and are not in this crop. The table has 15 printed columns and two heading levels: the spanner "Number of elections in which representation rights were won by unions" covers the four columns Total / AFL-CIO unions / Other national unions / Other local unions, and "Valid votes cast for unions" covers a second set of the same four column labels; the remaining headings are tall single headings repeated in both levels. Indentation in the stub (not preserved in the grid): "Public Administration" is a subtotal of the four preceding Executive/Justice/Administration/National Security lines; "Unclassified Establishments" and "Total, all industrial groups" are set at the same indent and are separated by double rules. The last heading is printed as "Eligible employ- ees in units choos- ing rep- resentati on"; the final broken word was joined as "representation". All values are as printed; the digit grouping is inconsistent in the total row (202,916 and 83,505 carry commas while 4276 and 5866 do not) but is transcribed as shown. No arithmetic inconsistencies were found: the Public Administration subtotal and the grand total both agree with their components as printed.