What the series proves

1899–1992 · the count read whole

First, the transformation had a measurable shape. The works-by-works ladder in the leg this addendum joins could date the great shops one by one; the federal count now frames them: a town of 56 small establishments and 1,761 wage earners in 1899 became, by 1919, a city whose 32 establishments averaged 158 wage earners each, and whose invested capital had multiplied six and a half times in twenty years. The consolidation happened between the censuses of 1914 and 1919 — establishments down twenty, wage earners up half — and it never reversed: no later census ever counts as many as the 52 establishments of 1914 in the old sense, and every later table shows the same structure of a few large works carrying the city's totals.

Second, the Bureau's refusals are a statistic of their own. Three times across the century the federal count could not publish Mishawaka without disclosing a single company: the whole city suppressed in 1935 and 1937, the money columns dashed in 1939, the industry detail of 1967 reduced to one small metalworking group. A diversified factory town does not trigger that rule; a city whose manufacturing is one works with satellites does, repeatedly, across four decades and three reorganizations of the census. The suppression footnote — "withheld to avoid disclosing data for individual establishments" — is the federal government's own characterization of Mishawaka's industrial structure, printed in its own volumes, and it corroborates from the outside what the Ball Band record shows from inside.

Third, hands and dollars part company after 1963. Employment peaks at 10,245 in 1963 and falls, census by census, to 5,300 in 1987 before recovering slightly to 5,600 in 1992. Shipments move the other way in every interval the tables print both: $214.1 million in 1967, $368.8 million in 1977, $416.4 million in 1982, $473.7 million in 1987, $873.7 million in 1992. These are nominal dollars across three inflationary decades, and the tables do not deflate them; but the divergence itself — the city's factories employing half as many people at the end of the series as at its peak while shipping several times the value — is the federal silhouette of the story the leg's later articles tell in works and names: the rubber age's mass employment gone, a smaller, capital-heavier industry, anchored by vehicle and machinery production, in its place. The horsepower column tells the same story's first chapter from the other end: 7,610 primary horsepower in 1909, 16,820 in 1929 — and then the city tables stop counting power altogether, just as the city's works stopped making their own and the purchased current of the Twin Branch age took the question out of the census's hands.

Sources:

1. The city rows printed in the five articles above, each from its volume, table and page there cited; the withholding rule as printed in the Biennial Census of Manufactures, 1937, cities index, and in the footnotes of the 1939 and 1967 tables there cited.