Valid Votes Cast in Representation Elections, by Final Results of Election, in Cases Closed, Fiscal Year 1982

p. 307. 14 columns: stub, "Total valid votes cast", then six columns under "Valid votes cast in elections won" and six under "Valid votes cast in elections lost". Section line: "A All representation elections" is printed centred across the value columns (not in the stub); I placed it in the stub cell with the other cells empty. A period after the "A" would be normal for this series but is not legible in the scan, so I transcribed it without one. Panel "B" (if any) falls outside this crop — the crop ends after "Total representation elections" with no further body, footnote text or rule. The caption carries a footnote marker 1, but no footnote text is printed inside the crop, so `footnotes` is empty. Arithmetic check (not used to alter any cell): every column adds correctly — the four 1-union rows sum to the "1-union elections" line, the nine 2-union rows to "2-union elections", the six 3-union rows to "3 (or more)-union elections", and those three subtotals to "Total representation elections", in all 13 value columns. Each row also satisfies won-total + won-no-union + lost-total + lost-no-union = total valid votes cast, and the union sub-columns sum to their "Total" within each panel. No misprints detected. Scan artefacts: the dotted i's in "unions"/"elections" are largely lost in this 1-bit scan and the leader dots before some 2-union rows are faint; neither affects the values. "v" (versus) in the row labels is transcribed as printed.

p. 308. Continuation page: no column headings are printed on this crop, so header_rows is empty. The grid has 14 columns (stub plus 13 numeric columns), counted from the vertical column rules; many cells are legitimately blank because each row reports votes only in the columns for the union(s) actually on the ballot. Crop boundaries: section B ("Elections in RC cases") begins mid-table, so the rows of section A and any part of section B above "AFL-CIO" are not on this image. At the bottom, section C ("Elections in RM cases") is cut off immediately after its "1-union elections" subtotal row — the 2-union, 3-or-more-union and "Total RM elections" rows continue on the following page. Arithmetic check: every one of the 13 numeric columns reconciles exactly in all four subtotal/total rows ("1-union elections", "2-union elections", "3 (or more)-union elections", "Total RC elections", and section C's "1-union elections"). No misprints detected. Scan quality: this is a 1-bit scan that drops small dots — the letter "i" prints dotless throughout ("unions", "national", "elections"). Consequently the periods that almost certainly follow the section letters ("B.", "C.") and the abbreviation "v." in the row labels are not visible on the image; I transcribed only the characters I can actually see. The hyphen in "AFL-CIO" may be an en dash in the original. Stray mark: in the section C row "Other local unions" there is an isolated dot between the columns holding 69 (the 10th numeric column region) and 69 (12th numeric column) — i.e. in or near the 11th numeric column. It is a speck/period artifact, not a digit, and the column subtotal (11 in the 11th column of "1-union elections", accounted for entirely by the "Other national unions" row) confirms the cell is empty; I transcribed it as blank.

p. 309. Continuation page: the body opens in the middle of the RM-cases panel (its 1-union-election rows and the start of the 2-union block are on the preceding page), so the first four rows carry no section line of their own. Column placement of the "Teamsters v national" row (4th body row) is the one doubtful point. Its three isolated zeros (one in the "won" half, two in the "lost" half) are printed under "Other local unions", not under "Other national unions" where the row label would lead a reader to expect them; I transcribed them in the columns they are physically printed in. Arithmetic cannot settle it, since the zeros add nothing either way: both the 2-union subtotal (other national 0 / other local 1,584 won; 0 / 2 lost) and the row total (55 = 32 + 23 + 0 + 0) check out under either placement. This looks like a typesetting slip in the original. Every other total and subtotal on the page cross-foots exactly, in both directions, which is the main basis for my confidence in the rest of the grid. The scan is a low-contrast 1-bit image in which small punctuation is unreliable. No period is visible after the "v" in the versus-style row labels, nor after the section letter in "D Elections in RD cases"; the inter-word gaps are wide enough that periods may well be present in the original print but dropped by the scan. I transcribed only what is legible. Likewise the dots over "i" are largely missing in the headings ("valıd votes", "Votes for unıons"), which I rendered with normal spelling. Several cells hold a single stray dot or a pair of dots rather than a number, and are transcribed as empty: "AFL-CIO v local" (RM) at won/Other national unions; the RD "AFL-CIO" row at won/Teamsters, won/Other national unions and won/Other local unions; the RD "Teamsters" row at won/Other local unions; the RD "Other national unions" row at won/AFL-CIO unions; and "Teamsters v local" at won/AFL-CIO unions (a doubled dot). Stray specks also sit in the stub beside "2-union elections" (RD), "3 (or more)-union elections" and "AFL-CIO v AFL-CIO v Teamsters"; these are not part of any label. The "D Elections in RD cases" line is a centred section banner spanning the whole body, carried in the stub with empty value cells.