NLRB Stats Archive › Annual Reports › FY1951 › Number of decertification elections and number of votes cast for participating unions during fiscal year 1951
| Participating unions | Employees eligible to vote | Percent casting valid votes | Total valid votes cast | Valid votes cast for— | Employees in units choosing representation | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Participating unions | Employees eligible to vote | Percent casting valid votes | Total valid votes cast | A. F. of L. affiliates | C. I. O. affiliates | Unaffiliated unions | No union | Employees in units choosing representation | ||||
| Participating unions | Employees eligible to vote | Percent casting valid votes | Total valid votes cast | Number | Percent | Number | Percent | Number | Percent | Number | Percent | Employees in units choosing representation |
| A. ELECTIONS | ||||||||||||
| Total | 93 | 15 | 8 | 4 | 66 | |||||||
| A. F. of L. affiliates | 50 | 15 | 35 | |||||||||
| C. I. O. affiliates | 23 | 8 | 15 | |||||||||
| Unaffiliated unions | 19 | 4 | 15 | |||||||||
| A. F. of L. affiliates—C. I. O. affiliates | 1 | 0 | 0 | 1 | ||||||||
| B. ELIGIBLE VOTERS AND VALID VOTES CAST | ||||||||||||
| Total | 6,111 | 87.5 | 5,350 | 1,090 | 20.4 | 940 | 17.6 | 366 | 6.8 | 2,954 | 55.2 | 2,632 |
| A. F. of L. affiliates | 2,831 | 87.3 | 2,472 | 1,078 | 43.6 | 1,394 | 56.4 | 1,225 | ||||
| C. I. O. affiliates | 1,725 | 84.6 | 1,460 | 938 | 64.2 | 522 | 35.8 | 1,187 | ||||
| Unaffiliated unions | 1,420 | 92.9 | 1,319 | 366 | 27.7 | 953 | 72.3 | 270 | ||||
| A. F. of L. affiliates—C. I. O. affiliates | 135 | 73.3 | 99 | 12 | 12.1 | 2 | 2.0 | 85 | 85.9 | 0 | ||
23 of 36 figures sit in a row or column sum that works out. 3 cells break a sum: 'Total', column 4: prints 8, the subtotals in columns 5 to 11 add to 4; 'C. I. O. affiliates', column 4: prints 8, the subtotals in columns 5 to 11 add to 0; 'Total', column 13: prints 2,632, rows 'A. F. of L. affiliat' to 'A. F. of L. affiliat' add to 2,682.
p. 318. STRUCTURE: the crop holds one table (Table 13) printed as two vertically stacked panels with DIFFERENT column structures, each with its own centered sub-caption and its own heading block. Panel A ("A. ELECTIONS") has a stub plus 5 value columns; panel B ("B. ELIGIBLE VOTERS AND VALID VOTES CAST") has a stub plus 12 value columns. Because the grid must be rectangular with one set of header rows, the grid is laid out on panel B's 13 columns (the wider panel), and header_rows reproduces panel B's three printed heading levels only. The two centered sub-captions are transcribed as label-only rows so the panel boundary is recoverable; they are printed centered above each heading block, not in the stub column. PANEL A COLUMN MAPPING (its headings are NOT in header_rows): panel A's five value columns are placed, in printed order, in grid columns 1-5. Panel A's own printed headings are, in order: col 1 "Number of elections"; then a spanner "Elections won by—" over col 2 "A. F. of L. affiliates", col 3 "C. I. O. affiliates", col 4 "Unaffiliated unions", col 5 "No union". Its stub heading is "Participating unions". So in the panel A rows, e.g. Total 93 = number of elections, 15 = won by A. F. of L. affiliates, 8 = won by C. I. O. affiliates, 4 = won by unaffiliated unions, 66 = won by no union. Grid columns 6-12 are empty for every panel A row. DECIMAL POINTS: all percent columns are printed with a decimal point that the 1-bit scan renders faint or drops, so the page reads e.g. "87 5", "20 4". They are transcribed as 87.5, 20.4, etc. Each one is consistent with the counts in the same row (5,350/6,111 = 87.5; 1,090/5,350 = 20.4; 938/1,460 = 64.2; 85/99 = 85.9), so the reading is secure. SUSPECT CELL: panel B, row "Total", column "Employees in units choosing representation" reads 2,632 as printed, but the four component rows sum to 2,682 (1,225 + 1,187 + 270 + 0). Either the printed total is a misprint or the scan has turned an 8 into a 3; it is transcribed as printed. Every other total in the table adds correctly (eligible 6,111; valid votes 5,350; A. F. of L. 1,090; C. I. O. 940; unaffiliated 366; no union 2,954; elections 93). PUNCTUATION: the periods in the abbreviations "A. F. of L." and "C. I. O." drop out inconsistently in this scan (headings and some row labels show "A F of L", "C I O"); they are rendered in full here because the underlying print carries them. Likewise the sub-caption for panel B prints as "B ELIGIBLE VOTERS AND VALID VOTES CAST" with no visible period after the B; it is given as "B." to match "A. ELECTIONS". OTHER: dash rules (------) in cells are transcribed as "". Printed zeros are kept as 0 (panel A joint row, and panel B last column). A stray ink speck sits just left of "99" in panel B, row "A. F. of L. affiliates—C. I. O. affiliates", column "Total valid votes cast"; it is not a decimal point and is ignored. No footnotes are printed under the table.
Source data: U.S. NLRB Annual Reports (public domain). Reconstruction by NLRB Stats Archive. Read by claude-opus-5 (effort max, prompt extract_v1) from the 300-dpi scan; stitched by page. Updated October 2026.