Valid Votes Cast in Representation Elections, by Final Results of Election, in Cases Closed, Fiscal Year 1999

p. 143. 11 columns plus the stub. "Total votes for no union" is a tall heading spanning heading levels 2-3 within each of the two "Valid votes cast in elections won/lost" blocks. Dash-rule cells ("----", meaning not applicable) are rendered as "" per the transcription rules; note that the table distinguishes these from printed zeros, and that distinction is lost in the grid. Examples where the two appear side by side: the "AFL-CIO v. AFL-CIO v. Local" row prints 0 under "Other local unions" (elections won) but "----" under "Other national unions"; the "3 (or More) union elections" row prints "----" under both "Other national unions" and "Other local unions" in the elections-lost block, whereas the identical-valued row above it prints 0 under "Other local unions". Doubtful cell: row "2-union elections", column "Valid votes cast in elections lost / Votes for unions / AFL-CIO unions" is printed as "320" followed by a short trailing dash ("320-"), transcribed as printed. The dash appears to be a stray mark or broken rule rather than a sign or footnote marker: the five 2-union detail rows above sum to 274+0+46+0+0 = 320, and 39,545+320+0 = 36,336... (specifically 36,016+320+0 = 36,336) matches the printed "Total representation elections" value in that column. All other cross-checks are internally consistent: each detail row's four component columns sum to its "Total valid votes cast", each subtotal matches its detail rows, and "Total representation elections" equals the sum of the three subtotal rows in every column. The caption carries footnote marker 1 ("Fiscal Year 1999^1"), but no footnote text is printed below the table in this crop; the footnote presumably appears elsewhere on the page or on a later page. The body is labeled section "A. All representation elections", so a section B of the same table almost certainly continues on a following page outside this crop.

p. 144. The crop shows only panels B and C of the table; panel A (and the caption and any footnotes) are not on this image, so the table is continued from an earlier page. No footnote markers appear anywhere in the crop. Column structure: 12 columns — stub, "Total valid votes cast", then five columns under "Valid votes cast in elections won" (Votes for unions: Total / AFL-CIO unions / Other national unions / Other local unions, then "Total votes for no union"), and the same five under "Valid votes cast in elections lost". Dash rules ("----") appear in many cells and are transcribed as empty cells per instruction; printed zeros are kept as 0. Note that the subtotal and total rows print 0 where the detail rows above print dash rules (e.g. "1-union elections" in panel C), which is as printed. In panel C the row labels are vertically offset from the data rows because of the typesetting: the stub label "AFL-CIO" is broken across two lines ("AFL-" / "CIO....") and sits beside the first two ruled data lines, and the labels "1-union elections", "AFL-CIO v. AFL-CIO", "Local v. Local", "2-union elections" and "Total RM elections" each sit low relative to their ruled data box. I paired each label with the data box it belongs to by analogy with panel B's structure (member rows, then the n-union subtotal, then the grand total); the label-to-row pairing in panel C is therefore the least certain part of this transcription, though the numbers are internally consistent with that reading (1,140 = 1-union subtotal of the single AFL-CIO row; 472 + 32 = 504; 1,140 + 504 = 1,644). The panel-C label "2-union elections" is printed with a wide gap before "elections" and trailing leader dots on the following line; it is transcribed as one label.

p. 145. The crop shows only panel D ("D. Elections in RD cases") of Table 14; panels A-C are on earlier pages, and the footnote marker 1 is attached to the caption (caption omitted per instructions). The section line "D. Elections in RD cases" is printed as a centered banner spanning the full width below the headings; it is transcribed as a label-only row. Cells printed as the dash rule "----" are transcribed as empty strings; printed zeros are kept as 0. All printed figures are internally consistent (each row's total valid votes cast equals the sum of its four won/lost components, and the 1-union, 2-union and Total RD elections lines equal the sums of their member rows), so no misprints are suspected. No cells were illegible.