Final Outcome of Representation Elections in Cases Closed, Fiscal Year 1995

p. 163. CROP IS INCOMPLETE. The image ends in the middle of section B: the last fully printed row is "National v. Local", and the table clearly continues below the crop (the remaining 2-union rows, any 3-union rows, a "Total" for section B, and presumably further sections). No footnote texts are visible on this crop, so `footnotes` is empty even though footnote markers 1 (in the caption) and 2 (on the "Total elections" heading) are printed; the marker-1 and marker-2 texts must be on the part of the page not captured here. COLUMN COUNT / STRUCTURE. 14 columns: stub, "Total elections^2", five columns under "Elections won by unions" (Percent won / Total won / AFL-CIO unions / Other national unions / Other local unions), "Elections in which no representative chosen", five columns under "Employees eligible to vote" (Total / In elections won / then three under "In units won by": AFL-CIO unions / Other national unions / Other local unions), and a final standalone column "In elections where no representative chosen". The last column is set off by its own rule and the "Employees eligible to vote" spanner rule stops before it, so I treated it as a tall independent heading rather than a sixth member of that group, although semantically it also counts employees. DECIMAL POINTS IN THE "PERCENT WON" COLUMN. This is a 1-bit scan and the decimal point drops out of many of these one-decimal percentages; they print as e.g. "50 0", "61 8", "46 3", "72 5", "89 7", "100 0", "47 4", "48 6", while others in the same column ("45.2", "83.3", "100.0", "51.2", "67.1", "71.4", "89.5") retain a clearly visible point. I transcribed every cell in this column with the decimal point. This is supported by arithmetic: each equals Total won / Total elections to one decimal (e.g. 44/88 = 50.0, 115/186 = 61.8, 1,525/3,296 = 46.3, 35/39 = 89.7, 1,611/3,399 = 47.4). DASHES. The em-dash rules used throughout for "not applicable" are rendered as "". Note the table distinguishes these from printed zeros: detail rows mostly use the dash while subtotal rows use "0" (compare "AFL-CIO v. AFL-CIO v. Local", which has a dash under Other national unions, with the identical-valued subtotal "3 (or more)-union elections", which prints 0 there). DOUBTFUL CELLS. Row "AFL-CIO v. AFL-CIO v. Local", column "In elections won": the scan shows a stray mark after the digits, reading roughly "268." — I transcribed "268", which is confirmed by the identical subtotal row below it. Section-B heading: the period after "B" is lost in the scan (it prints as "B Elections in RC cases"); I wrote "B. Elections in RC cases" to match the clearly printed "A." above. The AFL-CIO hyphen is a dash-like glyph throughout; written as a plain hyphen. ARITHMETIC CHECKS ALL PASS, which raises confidence in the column assignments. Section A: the three union-type rows sum exactly to "1-union elections" in all 12 count columns (e.g. 186,842 + 6,420 + 12,591 = 205,853; 115,184 + 3,844 + 6,808 = 125,836); the five pairings sum exactly to "2-union elections" (e.g. 3,219 + 904 + 3,554 + 751 + 588 = 9,016); and 1-union + 2-union + 3-or-more sum exactly to "Total representation elections" in every column (e.g. 205,853 + 9,016 + 268 = 215,137; 5,783 + 2,390 + 268 = 8,441). Section B likewise: 165,841 + 5,726 + 11,235 = 182,802 and 103,247 + 3,342 + 6,252 = 112,841. No misprints detected.

p. 164. Continuation page of Table 13 (14 columns). The crop carries column headings, so they are transcribed; no footnote texts are printed on this crop, although footnote markers appear in the caption (1995^1) and in the heading "Total elections^2" — their texts must be on another part of the page. The crop is cut off at the bottom: section "D. Elections in RD cases" breaks off after the "AFL-CIO v. Local" row, with no "2-union elections" subtotal or "Total RD elections" row, so the table continues beyond this image. Faint/dropped periods in this 1-bit scan: the "Percent won" cells for "Local v. Local", "3 (or more)-union elections" and "AFL-CIO v. Local" read as "100 0" with no visible decimal point; they are transcribed as "100.0" to match the clearly printed "100.0" in the other rows. Likewise the section line reads "C Elections in RM cases" with the period after "C" not visible; it is transcribed as "C. Elections in RM cases" to match the clearly printed "D. Elections in RD cases". Many cells hold an em-dash rule ("——") rather than a number; per instruction these are transcribed as "". In the RM-section row "AFL-CIO v. AFL-CIO" only one dash per half is clearly resolvable at this scan resolution (the other cell may hold a dash that dropped out, or be genuinely blank); either way both cells transcribe as "", so the grid is unaffected. Arithmetic cross-checks all reconcile as printed (e.g. Total RC elections: 1,273+48+135=1,456; 1,456+1,404=2,860; 65,925+2,797+7,641=76,363; 76,363+115,462=191,825; Total RM elections and the RD "1-union elections" subtotal likewise). No suspected misprints. The stub shows "AFL–CIO" set with an en-dash-like rule; transcribed with a plain hyphen as "AFL-CIO".

p. 165. This crop is the tail of Table 13 (a "—Continued" page): it carries the full three-level column heading but only the last two body rows of the RD-election panel ("2-union elections" and "Total RD elections"); the earlier rows of that panel are on the preceding page. In the "Percent won" column the decimal point is lost in the scan — the cells read as "50 0" and "29 3" with a gap where the point belongs. I transcribed them as 50.0 and 29.3, which is consistent with the printed counts (1/2 and 143/488) and with the one-decimal percent format used throughout this table. All other arithmetic in both rows is internally consistent as printed (136+1+6=143; 143+345=488; 8,976+192+790=9,958; 9,958+12,245=22,203; 91+2=93). The scan is slightly speckled and the headings show broken dot-accents (e.g. "unıons", "electıons"), but no digit is ambiguous.