The Weeks Report database


Peter B. Meyer (pbmeyer at econterms.net)
(formerly of Northwestern University Department of Economics,
now of U.S. Bureau of Labor Statistics, though this publication is not specifically authorized by those institutions)
Last updated June 19, 2004


This document describes the electronic form of report data set on wages drawn from Vol. XX ("Wages in Manufacturing") of the 1880 U.S. Census. It is called the Weeks report for Joseph D. Weeks, the special Census agent who ran the project. The data are reports of wages and job titles from a survey of manufacturing firms.

Thanks very much to Slil (Lynn) Siripong and Mercedes Delgado-Garcia for their superb assistance in the research project to build this database.

To cite this data set and its documentation here is a possible form:

Meyer, Peter B. (Ed.) 2004. The Weeks Report database, 4th edition. Available online: http://econterms.net/weeksreport/weeksdoc.htm And please at some stage send an email to pbmeyer (at) econterms.net letting us know you used it.


1. The data itself

The data set has about 100,000 observations of wages per job-year, mostly from 1860-1880, but extending overall from 1801 to 1884.

4th edition, from 19 June 2004, documentation and links to the data.
3rd edition, from 28 Oct 2003, in ASCII format. Has wage and other errors corrected.
3rd edition, from 28 Oct 2003, in Stata 7.0 format.
2nd edition, from 24 March 2002, in ASCII format. Has 102,537 observations.
2nd edition, from 24 March 2002 in Stata (ver 7.0) database format
1st edition, from 11/2001

2. Data fields (or columns or variables)

Following are the data columns and their explanations. Each observation has a wage and the other variables that are associated with that wage.

Field name
Contents
Firm #
 Possible values are in the range {100, 7999}. Each firm has been assigned a number. These numbers are grouped into industries. For example, the firms numbered from 4900-4999 are in the Belting industry.
Firm name
 Name of the firm, e.g. "Singer Manufacturing Company". May be akin to "An establishment in Ohio" for firms that did not wish to be identified.
Industry
 Text description from the Weeks report. See later table of industries.
City
 Blank if not known, otherwise is the name of the city in which the firm's main production occurs. May occasionally not be the city in which this particular worker works.
State
 Values are two-character abbreviations suitable for the year 2000. The state is known in almost all records (but one lists "New England", and another lists "Dakota Territory" which is categorized in this edition as South Dakota.)
Year firm was founded
 Left blank if not known.
Year
 Values are in the range 1820-1884. The survey data was collected in the early 1880s and was published in 1886. Data from before 1880 was reported retrospectively by firms from their records. Since (with one exception) only firms surviving to 1880 were surveyed, there could be survival biases in the data as a result. That is, it does not have a random sample of firms in earlier years, only those which survived until 1880.
Job title
 Text phrasing used in the Weeks report.
Standardized job title
 A content-oriented job title that is designed to group substantively similar jobs together. For example, in the blast furnace industry, "Top filler" and "Bottom filler" were often listed together so the standardized job for both is "Filler". The standardized job title is Printer for all of these: a Printer, an Apprentice (to a printer), a Helper to a printer, and an Overseer of printers.
Rank within firm
 Values are in the range {1,4}. Distinguishes records which represent various wages paid to workers with the same job title in the same firm in the same year. The highest wage gets a 1, and consecutively after that. The field is left blank in records which are already unique based on firm, year, and job title.
Represents group
 Values {0,1}. Most observations were summary statistics on a group of workers, usually the average. The value in this field is zero only when we know that exactly one worker is represented by this wage. The reason to track this information is that for some estimations researchers wish to put greater weight on the observations of undercounted workers, and in such cases they do not wish to over-weight the observations that are definitely of one worker, not an average of a group.
Wage
 A dollar figure. Represents the amount paid to a worker or an average, maximum, or minimum of pay to a class of workers with a certain job title.
Units of wage
 The wage figure may represent pay for a day of work (this is the most common), a week, a month, or the pay for some unit of output (a keg of nails, a dozen boots, one "heat" of a blast furnace). The most common unit is "day."
Overtime too?
 Value is 1, 0 or -1. Is 1 if the workers received either pay for overtime, or other benefits or pay not documented in the wage report. Examples are subsidized housing or sickness/injury support. Value is 0 if this is not known; -1 if worker definitely did not receive overtime or other pay besides that documented in the wage figures.
Piece rate
 Value is 0 or 1. Set to 0 if the worker was paid in time units. It is set to 1 if worker was paid according to measured output. Piece rate is always 1 if the "units of pay" is not in time units, but is also 1 if the original payment was by piece rate even if there was a later conversion to average daily income before the data was entered into the Weeks report.
Months worked this year
 A number from 0.1 to 12.0. This is a business cycle indicator. During the depression following the Panic of 1873 the average number of months worked fell noticeably.
Days worked per week
 A number from 0.1 to 7.0. A business cycle indicator. 6 days was standard. The average number fell in the depression after the Panic of 1873.
Hours per day
 A number varying from 8 to 15.5. This is a business cycle indicator. 10 hours and 12 hours were common standards.
Strikes
 Possible values {1/0/-1}: 1 if the firm is known to have had strikes, 0 if not known, -1 if the firm is known not to have had strikes.
Particular strike year
 If the year of a particularly significant strike is mentioned, it is recorded here.
Management won strike
 Values 0 or 1. The data in the Weeksreport come from the firm's management who may make clear whether they think they emerged victorious from the strike. If so there is a 1 in this field.
Is female
 Value is 0 or 1. This variable is 1 only when we know or impute that most workers represented in this group were female. In cases of doubt, it is zero.
Is child
 Value is 0 or 1. Takes value 1 for children (apparently meaning, in the context of the time, under 16). Value is 0 for adults or for cases that are uncertain.
Is black
 Value is 0 or 1. Takes value 1 for jobs whose workers are probably black.
Is Asian
 Value is 0 or 1. Takes value 1 for jobs whose workers have Asian ancestry.
Technology type
 Text listing attributes of the firm's technology. For example, for blast furnaces the type of fuel used: anthracite coal, bituminous coal, or charcoal.
Hot blast
 Value is 0 or 1. A technology attribute of iron blast furnaces. This field has a one only for blast furnaces that used the hot blast technology, not just the cold blast technology.
Labor-saving machinery?
 Values 0 or 1. Is 1 if the firm indicates adoption of a substantial amount of improved machinery during the survey period (whether or not it was intended to save labor).
Increase in labor efficiency?
 Values: {1,0,-1,-2} Is 1 if the firm reports improved "labor efficiency" in response apparently to a specific question by the surveyors. Is 0 if we don't know. Is -1 if firm reported that there was no increase. Is -2 if the firm reported a decrease.

Wages to total costs
 Value is in the set {blank, or 1-99}. The percentage of total costs of production spent on wage or labor expenses.
Wages to sales price
 Value is in the set {blank, or 1-99}. The percentage of the sales price spent on wage or labor expenses.
Product type
 Text list of the products produced at this firm, if known.
Steel?
 Value is 0 or 1. Value is 1 if this firm is a major maker or user of steel.
Paid in stuff
 Is 1 if the worker received a substantial part of documented earnings in a form other than cash, notably company store merchandise. Pay practices were evolving away from this toward payment in cash. This variable is not made to be 1 by the existence of some benefits like sick leave or subsidized rent.
Pay frequency, days
 Common values are 30, if pay was monthly; 7 for weekly; blank if not known.
Safe work?
 Values are in the set {-1,0,1,2,3} Value is -1 if there is evidence the work caused serious injuries or deaths, or if injuries were common. 0 if there is no evidence on the question. Values 1, 2, and 3 are for decreasing levels of dangerousness.
Reference to intemperate/lazy workers?
 Value is 0 or 1. Is 1 if the firm's comments include a statement that the workers drink too much, or are lazy.
Page
 Value is in {1,500}. Page in the published Weeks report.
This record rechecked?
 Values 0 or 1. Zero if no one except the person who entered the record has compared it to the original Weeks report.
Comments
 A text field. Notes especially if any information outside the Weeks report is included.
To do
 A text field; instructions to the database owner.


3. Industries represented

Industries in the Weeks report
Industry name
Number of firms
in data set
Number of
wage observations
Metal
work
 Iron blast furnaces
41
5496
Rolling-mills and nails
25
6762
Pins
2
148
Car-wheel founderies
8
421
Stove founderies
13
2650
General founderies
6
677
Hardware, cutlery, and edge tools
17
2357
Machinery
39
7623
Tin and sheet iron works
3
304
Agricultural implements
11
1352
Bells
1
70
Bridgebuilding
2
140
Agricultural
and
forestry
products
processing
 Canning
3
121
Cigars/tobacco
22
2048
Flour and grist mills
46
3500
Pork packing
3
89
Ice
2
326
Paper manufacture
36
6800
Sugar-refining
1
216
Breweries and distilleries
9
406
Textiles
 Carpets
3
742
Cotton manufacture
37
13596
Hemp and jute manufacture
1
160
Silk
4
1159
Wool
36
10385
Tanneries
22
2620
Hats
3
661
Clothing
5
605
Boots and shoes
13
2278
Belting
4
151
Wood work
 Cooperage
4
93
Furniture
41
5807
Saw- and planing-mills
52
6835
Ship-carpentry
4
437
Pianos and organs
4
1142
Carriage and wagon works
26
4411
Car-works
6
476
Mining,
minerals,
and
other
materials
 Iron mining
6
262
Brickmaking
10
1638
Coal mining
8
624
Copper mining
4
228
Silver mining
2
96
Stone quarrying
11
486
Powder (explosive)
2
23
Paints and white lead
5
321
Marble
3
299
Glass
17
5191
Gas and gas coke
6
303
Pottery and earthen-ware
23
1793


4. Descriptive statistics of Weeks report data 

Total observations:    104,413

In the averages below, piece rate wages, and wages reported by the week,
month, or year were left out.

Average wage of men:      $1.84 per day  (87362 observations)
Average wage of women:    $1.13 per day  (868 observations)
Average wage of boys:     $.67 per day   (2231 observations)
Average wage of girls:    $.76 per day   (1543 observations)

Earliest year observed:       1801
Median year of observation:   1873
Latest year observed:         1884

5. Job titles and content

This is a partial breakdown of which industries had which jobs:

Blast furnace workers -- Bloomary foreman
Hardware, cutlery, and edge tools -- Dropsman
Mining workers -- Colliers
Rolling mill workers -- Blast engineer, Bottom filler, Cinderman, Founder, Gutterman, Keeper, Molder, Reheater, Roller, Rougher, Top filler
Stone-cutting workers -- Setter, Stone-cutter
Glass workers-- Blower, or glass blower. Each glass blower worked with a mold boy and a cleaner-off. "The mold boy operates the molds into which the glass is blown, and the claner-off removes the particles of glass that adhere to the blower's rod." (Ashworth, 1915, p. 30, cited by Clawson (1980), p, 95)


Job contents:

Bloomary foreman, bloomer, or bloomsman -- operator of a bloomary, which was a charcoal-fueled furnace in which iron ore was converted ("reduced", chemically) directly into bar iron and liquid slag. -- Gordon (1996), p. 308.

Colliers -- mined coal, and/or made charcoal from wood. (Gordon, 1996, pp. 34-42)

Coopers -- Made barrels, buckets, and churns, usually from wood slats with edges angled so they would fit together in a cylinder, with wrought iron bands around them holding them together.

Dropsman -- This could have been a worker doing drop-forging, which is the rapid pressing-together of dies around a heated workpiece of metal, to form the metal into a certain shape.

Founder -- Person in charge of a blast furnace, or operator of a foundry. (Gordon, 1996, p. 309)

Keeper -- Artisan in charge of work done around the bottom part of a blast furnace. Usually the keeper tapped the furnace. (Gordon, 1996, p. 309)

Laborers -- the largest category. Mostly men.

Muck roller -- A rolling mill employee "who takes the bloom from the squeezer and throws it into the rollers." (Davis, 1922, chapter VII). A bloom is a blob of solidifying iron. A squeezer is a cylindrical thing that presses it into a little bit of shape and lets slag fall off. The rollers will press the bloom into a shape.

Reheaters -- in rolling mills -- were a new category around 1870, according to Montgomery (1987). Earliest ones in the Weeks report here were from 1868.
Roller -- top-ranked of the rolling mill employees. A roller ran a set of rolls, aided by roughers and various hooker-type aides.

Top filler -- Loaded a blast furnace with its inputs which were principally iron ore and coal, along with limestone and other helping components. The activity of loading was called charging the furnace. The inputs went into a vertical brick furnace "stack" which went down to the furnace's hottest point, where the iron ore would melt, separating into liquid iron and most of the remainder was called slag. The limestone was present to act as a "flux", an alkaline component, to combine with acidic parts to draw them off into the slag.

Wheelwright -- makes wheels, usually of wood, with a metal tire (a rim on the outside). Worked in carriage-and-wagon-works or in own standalone shops.



6. Future work

The data is now organized in one massive table, in which every record has a wage observation and characteristics of that job, including attributes of the firm.
In the future I hope to reorganize the information into separate tables, making a relational database. One table would have the wage, job title, establishment number, and year. Another would have establishment numbers along with attributes of the establishment like its name, industry, year it was founded, and anything we know about its products, and methods. Another would have information about jobs, including perhaps a description of the content of the job.

Until that kind of reorganization is done it is awkward to squeeze information about firms and jobs into the data because such information is attached to one or several records rather than to every occurrence of that firm name or job. (If it were attached to every occurrence, the data file size would swell up.) It's especially awkward to add information from sources outside the Weeks report, but wouldn't be so awkward in a relational format where it could be cited or discussed in the one row about a job or firm to which it was relevant.



7. Bibliography

Ashworth, John H. 1915. The Helper and American Trade Unions. Baltimore: Johns Hopkins University Press.
Clawson, Dan. 1980. Bureaucracy and the Labor Process: The Transofmration of U.S. Industry, 1860-1920. Monthly Review Press.
Davis, James J. 1922, I think. The Iron Puddler: My life in the rolling mills and what came of it. online at http://www.gutenberg.net/etext98/tirnp10.txt
Gordon, Robert B. 1996. American Iron: 1607-1900. Johns Hopkins University Press.
Meyer, Peter B. 2001. Technological uncertainty and earnings dispersion. Doctoral dissertation, Northwestern University Department of Economics.
Montgomery, David. 1987. The fall of the house of labor. Cambridge University Press.
Nuwer, Michael. 1988. From batch to flow: production technology and work-force skills in the steel industry, 1880-1920. Technology and culture.
Weeks, Joseph D. (editor) 1884. Report on the Statistics of Wages in Manufacturing Industries, 1880 Census Vol. XX.