Size of Units in Representation Election Cases Closed, Fiscal Year 1979

p. 323. 15 columns. Decimal points: this is a 1-bit scan in which the periods have dropped out, so every percent prints as e.g. "22 4" / "100 0". They are transcribed with the decimal point ("22.4", "100.0"), as in the source typesetting; the same applies to the panel label, printed "A Certification elections (RC and RM)" and transcribed "A. Certification elections (RC and RM)". No other character was altered. Verification: every count column sums exactly to its printed total — eligible to vote 538,404; total elections 7,266; AFL-CIO 1,923; Teamsters 1,017; other national 224; other local 265; no representative 3,837. The "Percent of total" column reproduces election counts / 7,266, and "Cumulative percent of total" is the running sum of those rounded percents, matching every printed cumulative value. Blank cells: dash rules ("------") printed in a percent cell wherever the adjacent count is 0 are given as "" (rows 180 to 189, 190 to 199, 500 to 599, 800 to 999, 2,000 to 2,999, 3,000 to 9,999). The cumulative-percent cell of the "Total RC and RM elections" row holds leader dots and is "". Two cells in the "Under 10" row do not equal the quotient of the printed counts: Teamsters percent reads 37.3 (381/1,017 = 37.5) and other national unions percent reads 13.7 (30/224 = 13.4). Both are transcribed as printed; with these values each percent column sums to exactly 100.0, so the largest size class appears to have been adjusted in the source to force the column total. Other local unions "Under 10" (19.8 vs 53/265 = 20.0) shows the same adjustment. Crop: the table is cut at the bottom edge immediately below the "3,000 to 9,999" row — the rule under it is not shown, and the remaining size classes of panel A plus any further panels (e.g. decertification elections) and any footnotes continue beyond this image. No footnotes are visible on the crop.

p. 324. Grid is 15 columns: stub + 14 data columns, all ruled, all falling under the single printed spanner "B. Decertification elections (RD)". This is a continuation page, so the only heading printed is that spanner; the individual column headings (and the stub heading) are on the preceding page and were not reconstructed. The column sequence inferred from the data is: employees eligible, elections, percent, cumulative percent, then six (number, percent) pairs; this is not printed on the crop. Periods reproduce very faintly in this scan and read as gaps at the resolution available. Every percent cell prints as e.g. "100 0", "31 4"; these are transcribed with decimal points ("100.0", "31.4"). The cumulative column confirms this (31.4 + 19.4 = 50.8, ... , 99.6 + 0.4 = 100.0). For the same reason the spanner, printed as "B Decertification elections (RD)" with a gap after the B, is transcribed as "B. Decertification elections (RD)", and the footnote is given a terminal period. Two doubtful cells, transcribed as read rather than corrected: - row "10 to 19", column 6 (percent beside the 114 column) reads "20 0" -> 20.0, but 23/114 = 20.2, and the parallel cell in the last column is 20.2. The final digit may be a 2 degraded to look like 0. - row "30 to 39", column 10 (percent beside the 21 column) reads "14 0" -> 14.0, but 3/21 = 14.3. The final digit may be a 3. Every other figure cross-checks: columns 2, 5, 7, 9, 11 and 13 sum exactly to 777, 114, 39, 21, 20 and 583; 114+39+21+20+583 = 777; column 1 sums to 39,538; all other percents and the whole cumulative column reproduce exactly. Cells printed as a dash rule ("---------"), including column 4 of the "Total RD elections" row, are given as "". Leader dots after row labels dropped. The header-row cell "777" carries a faint stray mark to its left (ink speck, not a comma or footnote marker); value confirmed as 777 by the column sum.