Final Outcome of Representation Elections in Cases Closed, Fiscal Year 1979

p. 309. 16 columns (stub + 15). Column assignments were checked arithmetically: in the 1-union and 2-union blocks every component row sums to its printed total in all 15 columns, and in each data row "In elections won" + "In elections where no representative chosen" = "Total" employees, so the placement of figures within the two wide groups is well supported. Percent won column: the decimal points are not resolvable in this 1-bit scan, so the cells read "42 6", "100 0", "0 0", etc. They are transcribed with the point ("42.6", "100.0", "0.0"); the one-decimal reading is confirmed by Total won / Total elections (e.g. 1,894/4,445 = 42.6, 290/375 = 77.3, 2/3 = 66.7). Nothing was rounded or recomputed. Header structure: "In units won by" appears to be printed over only the AFL-CIO unions, Teamsters and Other national unions columns (its centering matches a three-column span), with the following "Other local unions" column headed on its own at the same height; its figures nevertheless clearly belong to the "In units won by" breakdown (e.g. 93,459 + 25,910 + 19,460 + 10,818 = 149,647 in the 1-union elections row), so a reader may treat it as the fourth member of that group. "Total elections^2" carries a superscript 2; there is also a small raised dot after "Total" on the first line of that heading which is probably an ink speck or a damaged marker. A small stray squiggle sits in the lower right of the last column's heading box. The text of footnote 2 is not inside the crop, so no footnotes are returned. Section line: "A. All representation elections" is printed centred across the data columns (not in the stub); the period after the "A" is faint in the scan. It is recorded as a label-only row. Row-label periods: a period after the first "v" is visible only in "AFL-CIO v. AFL-CIO v AFL-CIO"; elsewhere the "v" shows no period. Periods are faint throughout this scan, so the original may print "v." in every label. Doubtful cells: - "AFL-CIO v national v local", Employees eligible to vote / In units won by / Other local unions: no figure is legible (blank or dash rule), although the row's own convention and the corresponding "Elections won by unions" cell (printed 0) suggest a 0 belongs there. Transcribed as "". - "AFL-CIO v local v local", Employees eligible to vote / In units won by / Other national unions: printed "0" even though the matching "Elections won by unions / Other national unions" cell is a dash rule; transcribed as printed (19,000 + 383 = 19,383 already balances the row). - "AFL-CIO v Teamsters v national": a small stray mark sits just left of the "0" in Elections won by unions / AFL-CIO unions; treated as a speck, cell read as "0". Crop: the table continues past the bottom edge — the total row for the 3-or-more-union group and any further panels (B, C, ...) of this table are on the following page. The caption and any page furniture are above/below the crop and are not included.

p. 310. The crop is a continuation page: no column headings are printed on it, so header_rows is empty. The grid has 16 cells per row (stub + 15 printed numeric columns); column rules confirm the count on the "Total representation elections" and "Total RC elections" rows, which fill every column. DECIMAL POINTS: in column 2 (the percent-won column) the scan drops the printed decimal point, so the cells image as "84 2", "45 0", "100 0", "0 0", etc. I have written them as 84.2, 45.0, 100.0, 0.0 and so on. Every one of them reproduces the ratio of column 3 to column 1 (e.g. 16/19 = 84.2, 51/81 = 63.0, 3,363/7,026 = 47.9), so the one-decimal reading is certain even though the point itself is not resolvable in the raster. No other column is affected. SUSPECT CELLS (transcribed as I read them, not corrected): 1. Row "AFL-CIO v national", column 9 (total eligible/valid votes): I read 7,127. That cell is inconsistent with the rest of the table — the row's own split is 3,127 + 4,543 = 7,670, and the column-9 entries of the ten 2-union rows sum to 49,451 against the printed "2-union elections" total of 49,994, a shortfall of exactly 543 = 7,670 - 7,127. Columns 10 and 15 of this row both reconcile with their subtotals, so the defect is in this one cell; the printed figure is probably 7,670 and either the scan or the original typesetting has it as 7,127. 2. Row "AFL-CIO v Teamsters", column 12 (Teamsters votes): I read 2,269. The row's union-vote cells give 1,140 + 2,269 = 3,409 against the printed row total of 3,309, and the column-12 entries of the 2-union rows sum to 3,733 against the printed subtotal of 3,633 — both discrepancies are exactly 100, pointing to a true value of 2,169. The glyph is genuinely ambiguous at this resolution between a 1 and a 2 in the hundreds place. Every other cell cross-foots: within each row, column 3 = columns 4-7 and column 1 = column 3 + column 8; column 10 = columns 11-14 and column 9 = column 10 + column 15; and the "1-union elections", "2-union elections", "3 (or more)-union elections" and "Total RC elections" lines all equal the sums of their member rows. MINOR MARKS: a stray speck sits just left of the 15 in column 10 of "AFL-CIO v AFL-CIO v AFL-CIO" (it images as "· 15"); it is dirt, not a footnote marker or a digit, and I transcribed the cell as 15. A similar speck sits inside "21,609" in column 14 of "Total representation elections". The section line reads "B" followed by a gap and "Elections in RC cases", centred across the full table width; the period after the B has dropped out of the scan the same way the decimal points have, and I have restored it. The caption, any part-A heading, and the running head "Appendix" / page number 29 are outside the transcription. No footnotes are printed on this page. Dash rules ("------") in empty cells are rendered as "".

p. 311. 16 columns: stub + "Total elections^2" + 6 under "Elections won by unions" + "Elections in which no representative chosen" + 6 under "Employees eligible to vote" + the final tall "In elections where no representative chosen". Decimal points: the scan has lost the decimal point in every value of the "Percent won" column, which prints as "24 7", "27 4", "16 7", "40 0", "25 6", "100 0", "27 5". I have restored the points ("24.7" etc.) the same way a comma-space artifact is repaired; each one is confirmed arithmetically by Total won / Total elections (37/150=24.7, 20/73=27.4, 1/6=16.7, 2/5=40.0, 60/234=25.6, 66/240=27.5). No digits were changed. The same tiny dot is lost after the section letter and after "v" in the four 2-union row labels; the image shows "C Elections in RM cases" and "AFL-CIO v AFL-CIO", and I have transcribed these as "C." and "v." for consistency with that reading. If a strictly literal rendering is wanted, drop those five periods and the seven decimal points. Header span, "In units won by": the printed rules put this heading over only three columns (AFL-CIO unions, Teamsters, Other national unions); the vertical rule to the left of the last "Other local unions" column rises a level higher, and that "Other local unions" heading is set as a tall box directly under "Employees eligible to vote". I have transcribed it as printed. Semantically "Other local unions" clearly belongs with the other three under "In units won by", so this may be a typesetting economy rather than an intended distinction. Footnote markers 1 (on the caption year) and 2 (on "Total elections") are printed, but no footnote text appears on this crop; the notes presumably sit on the first page of the table, so `footnotes` is empty. The crop is part C of a table continued from earlier pages; column headings are reprinted here, so they are transcribed. No row or value is cut off, and all 15 numeric columns are internally consistent: the four single-union rows sum to "1-union elections", and "1-union elections" + "2-union elections" sum to "Total RM elections", in every column. No suspected misprints. Dash rules ("------") in empty union cells are transcribed as "". Printed zeros are kept as "0".

p. 312. Structure: stub + 15 data columns (16 cells per list). This is a continuation page: the per-column headings sit on the preceding page and are NOT printed here, so nothing has been reconstructed for them. The one heading printed on this crop is the panel banner "D Elections in RD cases", set centered in the boxhead over all 15 data columns and above the rule that opens the body, with the stub area beside it blank; I therefore recorded it as a single heading level repeated across the 15 data columns with an empty stub cell. If your convention treats such a panel banner as a body section line instead, it is the line that would precede the "AFL-CIO" row. The character after "D" is not legible — the scan shows only a wide gap where a period would be — so it is transcribed as "D" with no period. Column 2 (the percent column) prints its decimal point too faintly for the scan to carry it: the cells read "23 9", "100 0", "25 0", etc. I restored the decimal point ("23.9", "100.0", "25.0"), which arithmetic confirms (107/447 = 23.9; 194/777 = 25.0; 13/14 = 92.9). The same period loss shows up at the end of both footnotes, which are transcribed without a closing period because none is visible. Cells printed as a dash rule ("------") are rendered as "". No footnote markers appear anywhere on the body of this crop; markers 1 and 2 belong to the caption/headings on the earlier page. One doubtful mark: a faint stray speck sits just above the "777" cell, on the double rule above the "Total RD elections" row. It does not read as a digit or a footnote marker and I treated it as scan dirt, so it is not transcribed. No misprints found. Every line reconciles as printed: columns 4-7 sum to column 3 and columns 3+8 = column 1 in every row; columns 11-14 sum to column 10 and columns 10+15 = column 9 in every row; and the three subtotal rows (1-union 743, 2-union 31, 3-or-more 3) sum to the printed Total RD elections line (777 / 194 / 583 / 39,538 / 17,450 / 22,088) in all 15 columns.