The Company That Cared: Reading Welfare Capitalism from Payroll Records, 1926
A state labor commission is evaluating a 1926 claim by the Granite Falls cotton mills that their welfare program improved workers' lives. The program includes a company hospital, a savings plan, a baseball league, a pension after 25 years, and an employee council; the company advertises a no-strike record. The commission has payroll and personnel data for 1926: 11,400 workers; average weekly hours 48, down from 54 in 1920; average weekly earnings $22.80, down from $24.10 in 1920; 3,600 workers enrolled in the pension plan, with 41 currently drawing pensions; the council has met nine times in three years and its minutes record no discussion of wages; annual turnover is 63%, and 71% of the workforce has been employed less than two years.
1. Was the welfare program a genuine improvement or a stabilization strategy? Cite the figures that support each reading and say which reading the data better support.
Read the worked answer
The company's claim has real support. Hours fell from 54 to 48, a genuine reduction in exposure; a hospital and a savings plan are material benefits. The earnings decline is less damaging than it looks: $24.10 to $22.80 is a 5.4% nominal drop across years when wholesale prices fell sharply after 1920, so real weekly earnings were probably flat to slightly up.
But the personnel data undercut the program as advertised. Turnover of 63% a year and 71% of workers with under two years' tenure mean the program retains almost nobody, and a pension that vests at 25 years is a promise most workers will never collect (3,600 enrolled, 41 drawing). The council met nine times in three years and never discussed wages: voice without bargaining.
Judgment: not cynical theater, but a stabilization strategy whose main product is the appearance of harmony, purchased cheaply. The test is whether it costs the firm anything it would rather keep, and wage rates are conspicuously untouched.
2. What additional evidence would separate the two interpretations, and which item would you seek first?
Read the worked answer
Four items would settle much of it.
- Pension funding: actuarial reserves versus pay-as-you-go from current revenue. An unfunded plan is a promise designed to be broken at the next downturn.
- Quit rates by department and stated reason. If quits cluster where hours were cut, the shorter week was experienced as lost pay, not leisure.
- A comparison mill: wages, turnover, and accident rates at a nearby firm without a welfare program. Welfare capitalism's significance depends entirely on the counterfactual.
- Grievance outcomes: whether the council ever reversed a foreman's decision. Nine meetings and no wage talk suggests not; one reversed firing would change the reading.
Weigh the comparison mill first, because the other three describe this firm's internal bargain. If the local labor market already delivered these conditions, the program is advertising. If it did not, it is a limited but real concession, likely pried loose by a tight labor market rather than by benevolence.
3. How would a historian using only company promotional material reach a different conclusion, and what does that reveal about source selection?
Read the worked answer
That historian would write a story of enlightened management: hospitals, pensions, recreation, a no-strike record, the language of partnership. Every claim would be self-reported, and conflict would be absent by construction. The result would not be a lie so much as an answer to a different question: what did the company want to be seen doing?
The lesson concerns source selection, not corporate mendacity. Promotional material is excellent evidence of corporate ideology, of what managers feared (turnover, unions, legislation), and of the audience they addressed. It is nearly worthless as evidence of workers' experience, which is why payroll, turnover, and grievance records are the necessary complement.
The transferable rule: ask what question a source can answer before asking whether it is biased. Notice the trap too: a historian with only the glossy pamphlet might still be right about the hospital existing, and wrong about almost everything that mattered.