Analysis of Methods of Disposition of Unfair Labor Practice Cases Closed, Fiscal Year 1996

p. 132. 18 columns: stub + 3 for "All C cases" (Number, Percent of total closed, Percent of total method) + 2 each (Number, Percent of total closed) for CA, CB, CC, CD, CE, CG, CP. Crop is cut off at the bottom: the table continues below the last transcribed row ("Compliance with", whose own indented sub-rows and the later method groups are not on this image), and no footnote texts are printed within the crop, so `footnotes` is empty even though markers 1 (on the caption, excluded) and 2 (in the body) are referenced. "Total number of cases closed", column "All C cases / Percent of total method": printed as 0.0. Transcribed as printed; this looks like a misprint or a placeholder, since that column is otherwise a within-method percentage (the method group headings "Agreement of the parties" and "Compliance with" carry 100.0 there). "Before issuance of complaint", column "CD cases / Number": the cell holds only a parenthesised raised 2 — a standalone footnote reference rather than a value. Transcribed as (^2). Cross-footing confirms the underlying count is zero (6,470+730+178+0+8+8+29 = 7,423), so the "2" is a marker, not the number 2. The paired "CD cases / Percent of total closed" cell holds a dash rule and is "". Dash rules (—) standing in for "no percentage" are transcribed as "" throughout; printed zeros are kept as 0. Verification: for every data row the seven case-type Numbers sum exactly to the "All C cases" Number (e.g. 22,703+5,942+482+194+35+25+104 = 29,485; 9,068+968+213+5+12+9+44 = 10,319; 636+44+8+1+0+0+2 = 691), which confirms column alignment across the 18 columns. Row labels "Stipulated decision" and "Consent decree" each occur twice (once under "After issuance of complaint, before opening of hearing" and once under "After hearing opened"); the printed indentation that distinguishes the hierarchy is not preserved in the stub text.

p. 133. 18 columns: stub + "All C cases" (Number, Percent of total closed, Percent of total method) + seven two-column panels (CA, CB, CC, CD, CE, CG, CP; Number and Percent of total closed each). Dash rules ("——"), which occupy most percent cells in the CC–CP panels, are transcribed as empty cells per instruction; printed zeros are kept as "0". CD-cases "Number" column: three cells appear to be printed as a parenthesized footnote reference, "(2)", standing in place of a value rather than as a count — rows "Before issuance of complaint" (under Withdrawal), "Before issuance of complaint" (under Dismissal), and "After issuance of complaint, before opening of hearing" (under Dismissal). The first two are legible as "(2)"; the third is the least certain and could conceivably be a plain "0" degraded by the scan. In all three cases the column arithmetic is consistent with a zero contribution (e.g., CA 7,002 + CB 1,624 + CC 156 + CE 13 + CG 7 + CP 25 = All C 8,827), so the cell carries no counted value either way. Note the footnote list is empty because no footnote text is printed on this crop; footnote 2 itself must be on the page where the table ends. The crop is a continuation page: the opening rows ("Administrative law judge's decision", "Board decision", "Circuit court of appeals decree", "Supreme Court action") are sub-items of a method group whose heading and total are on the preceding page — their "Percent of total method" values (4.2 + 61.9 + 33.9 + 0.0) sum to 100.0. The table is also cut off at the bottom: the last printed line is the "Contested" sub-row of "By Board decision", and the Dismissal components shown here sum to 9,075 (All C), 5,671 (CA) and 3,255 (CB) against printed totals of 9,081, 5,676 and 3,256, so further stage rows (and any footnotes) continue below the crop. Indentation levels are not preserved in the grid: "Adopting administrative law judge's decision (no exceptions filed)" and "Contested" are sub-items of the "Board decision" row above them (rows 3–4) and of the "By Board decision" row (rows 19–20) respectively; "Before issuance of complaint" through "After Board or court decision" are sub-items of "Withdrawal", and the parallel block is sub-items of "Dismissal". Minor stray specks from the scan sit next to some values (e.g., "0." in the CE column of "Supreme Court action", "7." in the CG column of "Withdrawal", ",0" in the CP column of the Dismissal "After hearing opened" row); these are read as plain digits. Arithmetic checks pass everywhere except where truncation explains the gap, which supports the digit readings. Source is the 1996 report, whose scan is lossy JBIG2 with a known digit-substitution risk, so isolated digits carry slightly more risk than usual.

p. 134. The crop is the tail of a table continued from the preceding page: column headings are printed, but the first two body rows ("By circuit court of appeals decree", "By Supreme Court action") are indented sub-entries whose parent heading appears on the earlier page. 18 columns: stub, then "All C cases" with three columns (Number, Percent of total closed, Percent of total method) and seven two-column panels (CA, CB, CC, CD, CE, CG, CP), each Number and Percent of total closed. Cells rendered "" hold an em-dash rule in the original. Footnote marker 2 is printed under the table but its referent does not appear anywhere on this crop (it attaches to a row or heading on the earlier page). Possible misprint worth flagging: in the "10(k) actions" row the "All C cases — Percent of total method" cell reads 0.0 even though the Number is 187; transcribed as printed. Digits in the 300-ppi scan are clean; 96.3 (CD, 10(k) actions row) and 187 were checked twice.