Final Outcome of Representation Elections in Cases Closed, Fiscal Year 1977

p. 309. Crop ends immediately below the "Total representation elections" row, so no footnote texts are visible; the list is empty even though two markers are referenced on the page (a "1" on the caption, excluded per instructions, and a raised marker on the "Total elections" column heading). That heading marker is a small, partly broken glyph; I read it as "2" (giving "Total elections^2"), but "3" cannot be fully ruled out at this scan quality. Column structure: 16 columns. "Elections won by unions" spans six columns (Percent won, Total won, AFL-CIO unions, Teamsters, Other national unions, Other local unions). "Employees eligible to vote" spans six columns (Total, In elections won, and four under "In units won by": AFL-CIO unions, Teamsters, Other national unions, Other local unions). "In units won by" is read as spanning all four of those sub-columns, including "Other local unions"; the spanning rule's right end is faint, but the arithmetic confirms it (e.g. 1-union elections: 110,185 + 31,602 + 14,719 + 10,911 = 167,417 = "In elections won"). Section line "A. All representation elections" is printed as a centered banner across the body with no values; the crop contains no panel B. Dash rules ("------") used as empty-cell fillers are transcribed as "". Printed zeros are kept as 0 and are distinguishable from the dash rules in every cell. Scan artifacts, all resolved by cross-footing: in "AFL-CIO v. Teamsters v. National" the "In elections won" value 465 and the "In units won by / AFL-CIO unions" value 0 carry stray ink that makes them look struck through; the readings 465 and 0 are confirmed since 0 + 0 + 465 = 465. In "Other national unions" and a few other rows the leader dots run close to the first digit, but no digit is obscured. Verification: every arithmetic relation in the table holds as transcribed — component union columns sum to "Total won"; "Total won" + "no representative chosen" = "Total elections"; the four "In units won by" columns sum to "In elections won"; "In elections won" + "In elections where no representative chosen" = "Total" employees; each percent matches "Total won" / "Total elections"; and the three subtotal rows (1-union 8,944, 2-union 511, 3-or-more 29) sum to the grand total 9,484 in every column. No printed value appears to be a misprint.

p. 310. Continuation page: no column headings are printed on this crop, so header_rows is empty. The grid has 15 value columns plus the stub; column structure was confirmed from the printed subtotals (cols 4-7 are a won-by breakdown summing to col 3, which plus col 8 gives col 1; cols 11-14 are the matching employee breakdown summing to col 10, which plus col 15 gives col 9). SUSPECTED MISPRINT: in "Total RC elections", column 5 is printed as 1,559, but its parts (1-union 1,089 + 2-union 69 + 3-or-more 1 = 1,159) and the row's own column 3 (2,359 + 1,159 + 236 + 302 = 4,056, as printed) both imply 1,159. The glyph after the comma reads as a 5 in the scan, so it is transcribed as printed; a reader should treat this cell as doubtful (either an original misprint for 1,159 or a 1/5 scan confusion). Every other cell in the table reconciles exactly with its printed subtotals and totals. The percent column prints one decimal throughout, but on this scan many of the decimal points are broken or lost (e.g. "44 8", "100 0", "75 0", "29 7"); they are transcribed with the decimal point restored, in line with the rule for the spaced thousands comma. Likewise the period after "v" in several row labels ("Teamsters v national", "Local v local", "AFL-CIO v AFL-CIO v. National", "AFL-CIO v. National v National", "Local v. Local v Local") is faint or dropped and has been written as "v." to match the rest of the table; capitalisation is kept as printed (lowercase "national"/"local" in the 2-union block, capitalised in the 3-or-more block). "B. Elections in RC cases" and "C. Elections in RM cases" are the panel/section lines spanning the full width and carry no values. The three long 3-or-more-union labels wrap over two or three printed lines and have been joined ("Team- sters" to "Teamsters"; "AFL-" + "CIO" to "AFL-CIO"). Dash rules ("------") are rendered as empty cells; printed zeros are kept as 0. A "See footnotes at end of table" line appears below the body, clipped by the bottom edge of the crop; there are no numbered footnotes on this page, so footnotes is empty. The running head "Appendix" and the page number 295 in the right margin are excluded. Panel C is cut off after its "1-union elections" subtotal, where the page ends.

p. 311. 16 printed columns; the crop continues Table 13 from the preceding page, so the body opens mid-section with the section line "C. Elections in RM Cases—Continued" (kept as a row because it carries a label; it is a within-body section head, not the caption's continuation line). Punctuation is badly inked throughout this 1-bit scan: many decimal points in the "Percent won" column print as a gap ("100 0", "40 0", "0 0", "31 3", "24 0") and the periods in the section-D row labels ("AFL-CIO v AFL-CIO", "D Elections in RD cases") are likewise invisible. Since every such value is a one-decimal percentage and the matching section-C labels do show the points, I have written the decimal points and label periods uniformly rather than reproducing the dropouts. Likewise the terminal periods of both footnotes are not visible and were supplied. Cells printed as dash rules (-------) are rendered "" per instruction; printed zeros are kept as 0. In section C the label is "AFL-CIO v. local" with a lowercase l, while the corresponding section-D label is "AFL-CIO v. Local" with a capital L — transcribed as printed in each case. A small stray mark sits immediately left of 733 in the last column ("In elections where no representative chosen") of the "Teamsters v. National" row; it is a speck or broken rule rather than a footnote marker, so the cell is given as 733. A similar faint tick appears above the rule between the "Elections won by unions" spanner and the header rule, and another in the blank band above "D. Elections in RD cases"; both are scan debris. No arithmetic was altered. All internal checks pass: component elections sum to each subtotal and total, won + no-representative = total elections, and the "In units won by" columns sum to "In elections won", which with the no-representative column sums to "Total" employees eligible, in every subtotal and in both "Total RM elections" and "Total RD elections".