The source: the 1871 manuscript census
The manuscript schedules for industrial establishments counted in the 1871 census constitute a uniquely valuable source for Canada. The wealth of detail in this source has become more accessible since the schedules were microfilmed by the National Archives as part of the whole 1871 manuscript census. See T.A. Hillman, Catalogue of Census Returns on Microfilm, 1666 1881 (Ottawa: Public Archives of Canada, 1981). The 1871 census was the first to be taken on a consistent basis for the provinces that formed the Dominion of Canada by that date. The census was directed by Dr Joseph Charles Taché, who was Canada's Deputy Minister of Agriculture from 1864 to 1888. The census procedures were very carefully planned and implemented "with the utmost accuracy possible" by "honest, intelligent, well-instructed and painstaking staff". The careful planning is evident in the "Manual Containing the Census Act and Instructions to Officers Employed in the Taking of the First Census of Canada, 1871" in the Canada Sessional Papers No. 64 (1871). Although similar information was collected in the Canadian censuses of 1881, 1891, 1901 and 1911, the manuscript schedules for industrial establishments in those years have not survived. Moreover, the 1871 schedules contain a great deal of information which was never published at the time. Not only are the 1871 census manuscripts the only systematic source surviving in such detail from the nineteenth century. They also mark a time of transition in industrial technology, business organization and work discipline.
Data for industrial activity in the four provinces of Canada during the twelve months preceding 1 April 1871 were recorded on the sixth of the nine schedules used in taking the census.
The nine schedules were as follows:
1) nominal return of the living, including details of names, gender, ages, family relationships and dwellings;
2) nominal return of the deaths during the past twelve months;
3) return of public institutions, real estate, vehicles and implements;
4) return of cultivated land, of field products and of plants and fruits;
5) return of livestock, animal products, homemade fabrics and furs;
6) return of industrial establishments;
7) return of products of the forest;
8) return of shipping and fisheries; and
9) return of mineral products.
As page and line references to the nominal returns had to be specified on each of Schedules 3, 4, 5, 7, and 8, it is possible now to interrelate the information in all these returns. Such cross-references are not possible with the industrial schedule, though enumerators were charged with asking every family they visited whether any member carried on any "industrial establishment" (p.131). The industrial schedule contained the only questions relating to dollar values in the whole census.
A manual of instructions was produced for the census enumerators in 1871. It is important to note the definitions of industrial activity that were expressed or implied. The enumerators were instructed to record all industry of any importance which was conducted in separate establishments or workshops. The instruction manual defined an industrial establishment as "a place where one or several people are employed in manufacturing, altering, making up or changing from one shape into another, materials for sale, use or consumption, quite irrespectively of the amount of capital employed or of the products turned out."
Specific industry types were expected to include "a lime kiln, a cheese factory, a brick-yard, a ship-building yard, a grinding-stone factory, a sulphuric acid manufactory, a saw-mill, a marble cutter's shed, a wheel factory, a pottery, a foundry, a meat-curing or "packing" establishment, an establishment to manufacture copper regulus, or purify plumbago, a cloth manufactory, a carding mill, a grist mill, a planing and dove-tailing mill, a sash factory, as well as shoe-making, harness-making, dress-making, tailor's, or blacksmith's shop, or carpenter's or joiner's shop, etc., etc."
It was stated that "all repairs, mending or custom work are understood to be industrial products; and are to be entered accordingly, by value, on the returns of industrial establishments". Thus the definition of industrial activity was considerably broader than it would be in the twentieth century. Enumerators were specifically instructed to record returns of Industrial Establishments in the geographical units -- District, Sub-District or Division -- in which they were found "and nowhere else. The principle is essential in every case. The production is attached to the locality".
The values to be stated for raw materials and output were not to be confused with any calculation of sales or profit. This provision applied to custom work such as tailoring or work done on toll such as that of some grist mills or carding mills:
“It matters not whether the raw material is in the ownership of the manufacturer or not, whether it is transformed on account of one or another person, whether the working is profitable or losing business; the information required is the result from the establishment, so far as requested. For instance, a saw mill may saw logs and other lumber for a great number of other persons; the working may cost more than the returns bring to the owner; but, nevertheless, the amount of raw material has changed form, and so much value has been added to it; and this is the fact to be recorded.”
Numbers of workers or "employés" were to include only those persons actually working in the industrial establishment: “...The number of people employed may be made up exclusively with members of the family of the proprietor; in other places the proprietor and family may not form part of the people employed.”
Though enumerators were invited to specify types, quantities and values of individual raw materials and products for each establishment, it was foreseen that usually only the aggregate values would be stated:
“In many cases the raw materials or articles manufactured are of such a multifarious character that they must be lumped together, and entered by the value.”
Definitions used in the Canadian census may be compared with those in the 1870 census of the United States. No minimum value of output was set, in contrast to the United States, where only establishments with at least $500 worth were included. The values of fixed capital and floating capital invested in the business were distinguished rather than merged and the number of working months was specified. Although the industrial workforce was subdivided by age and gender into men, women, boys (under 16 years) and girls (under 16 years), only "average numbers" of each employed during the census years were to be entered and the amount of wages paid to each group of worker was not separately stated.
The 1871 census enumerators recorded the following details for each establishment they found, relating to industrial operations in the twelve months preceding 1st April, 1871. These details were handwritten in numbered columns horizontally across the schedule forms, five establishments to the page.
 |
Facsimile of Manuscript Schedule in English |
1. Kind of Industrial Establishment, Name of Proprietor or Company, and other such information / Genres d'établissements industriels, noms des propriétaires ou des compagnies et autres renseignements.
2. Fixed Capital invested in $ / Capital fixe, en piastres.
3. Floating Capital employed in $ / Capital flottant, en piastres.
4. Number of working months in the year / Nombre de mois de travail dans l'année.
5 - 8. Average Number of People employed / Moyenne du nombre de personnes employées:
5. Male over 16 years / Hommes au-dessus de 16 ans;
6. Female over 16 years / Femmes au-dessus de 16 ans;
7. Boys under 16 years / Garçons au-dessous de 16 ans;
8. Girls under 16 years / Filles au-dessous de 16 ans.
9. Aggregate amount of Yearly Wages in $ / Montant collectif des gages durant l'année en piastres.
10. Moving Power: Kind / Force motrice: Nature.
11. Moving Power: Nominal force / Force motrice: Force nominale (stated in units of horse power)
12 - 14. Raw Materials / Matières brutes:
12. Kind / Espèces; (units of measurement often stated here as well)
13. Quantities / Quantités;
14. Aggregate Value in $ / Montant de la valeur en piastres;
15 - 17. Products / Produits:
15 Kind / Espèces; (units of measurement often stated here as well)
16. Quantities / Quantités;
17. Aggregate Value in $ / Montant de la valeur en piastres;
18. Remarks / Remarques (this space was sometimes used for additional information, perhaps an address, or an explanation of special difficulties encountered by the enumerator).
Each schedule was headed with details of the relevant Province, Census District, Census Sub-District and Census Enumerator's Division, including both the place names and the number and letter codes prescribed by the census organizers.
Only a very limited amount of the information collected in Schedule 6 of the 1871 census was published in the official census volumes of the 1870s. The published statistics were organized primarily by various industrial types which were defined pragmatically rather than systematically and then listed alphabetically. For each industry type, whether as specialized as whip making or as ubiquitous as blacksmithing, figures were published for numbers of establishments, hands employed, yearly wages, value of raw materials and value of products in each Census District.
No industrial data at all were published for smaller areal units such as the Census Sub-Districts, either as summaries of total industrial activity or for specific types of industry. Thus the only 1871 industrial information published for urban centres was for the six cities, the boundaries of which exactly coincided with those of one or more census districts. These were Montreal, Toronto, Hamilton, Ottawa, London and Kingston. None of the information collected on the industrial use of inanimate power was released. Furthermore, the published totals understated the real extent and value of industrial activity as these can now be reconstituted from the manuscript census schedules.
Back to top
The CANIND71 project has made machine-readable the full data for more than 45,000 industrial firms that were counted in Canada's first national census in 1871. For each establishment, there are up to 125 variables. The firms were located in 206 census districts and 1701 census sub-districts in Ontario, Quebec, New Brunswick and Nova Scotia. The CANIND71 project methodology was designed to make all this information accessible in systematic, standardized and readily retrievable format. The project team prepared the database and documentation to the highest standards of accuracy, clarity and consistency so that they can serve the needs of all users.
In this section, we explain the procedures that we used in entering and editing the information from manuscript schedules to the computer record. We define each of the variables in the CANIND71 database and discuss the main problems we faced in standardizing and coding data for each variable.
Throughout the project, the goal has been to make possible both the retrieval of information for individual firms and the orderly aggregation of data according to location and/or industry type and by various measures of size and significance. Thus the CANIND71 database supports research at all levels, from the particular enterprise in its context of place and industry type to the most generalized abstractions for whole provinces. In converting a major body of historical data into electronic format, we had to make and implement several critical decisions. These decisions were:
- to include all records rather than just a sample;
- to transcribe in the natural language used by the enumerators all information written on the original schedules;
- to add standardized codes to each record according to systems of locational references and industrial classification to help with data processing;
- to prepare the CANIND71 database and its documentation in order to facilitate accessibility in a variety of hardware and software environments. For more discussion of project procedures and an evaluation of the quality of the data, see Creating CANIND71: Procedures for Making the 1871 Census Machine-Readable (1989), #4 in the CANIND71 series of research reports.
CANIND71 Database Variables
The database comprises variables of two main types: the basic ones obtained by direct transcription of information from the microfilmed manuscript schedules, and the derived variables for which calculations or inferences were made from the basic variables.
Alphabetical list of variables.
Basic variables are explained first in relation to the format of the original schedules. Variable names used in the CANIND71 database are printed in bold capitals. We also present an alphabetically ordered explanation of all variable names.
1. Kind of Industrial Establishment, Name of Proprietor or Company, and other such information / Genres d'établissements industriels, noms des propriétaires ou des compagnies et autres renseignements. This information is contained in the PROPRIOR and TYPEEST variables; in a small proportion of records, lengthy details that could not be fitted in the available space were placed in the COMMENTS field.
2. Fixed Capital invested in $ / Capital fixe, en piastres - FIXCAP
3. Floating Capital employed in $ / Capital flottant, en piastres - FLOCAP
4. Number of working months in the year / Nombre de mois de travail dans l'année - MONTH
5 - 8. Average Number of People employed / Moyenne du nombre de personnes employées:
5. Male over 16 years / Hommes au-dessus de 16 ans - EMPMEN
6. Female over 16 years / Femmes au-dessus de 16 ans - EMPWOM
7. Boys under 16 years / Garçons au-dessous de 16 ans - EMPBOY
8. Girls under 16 years / Filles au-dessous de 16 ans - EMPGIRL
9. Aggregate amount of Yearly Wages in $ / Montant collectif des gages durant l'année en piastres - WAGES
10. Moving Power: Kind / Force motrice: Nature - TYPEPOW
11. Moving Power: Nominal force / Force motrice: Force nominale (stated in units of horse power) - FORCE
12 - 14. Raw Materials / Matières brutes: with provision for up to 12 different raw materials per establishment:
12. Kind / Espèces; (units of measurement often stated here as well) - RAWMAT1 ... RAWMAT12
13. Quantities / Quantités - RQUANT1 ... RQUANT12 with units of measurement as RUNIT1 ... RUNIT12
14. Aggregate Value in $ / Montant de la valeur en piastres - RVALUE1 ... RVALUE12
15 - 17. 1Products / Produits: with provision for up to 12 different products per establishment:
15. Kind / Espèces; (units of measurement often stated here as well) - PROD1 ... PROD12
16. Quantities / Quantités - PQUANT1 ... PQUANT12 with units of measurement as PUNIT1 ... PUNIT12
17. Aggregate Value in $ / Montant de la valeur en piastres - PVALUE1 ... PVALUE12
8. Remarks / Remarques: This space was sometimes used for additional information such as an address or an explanation of special difficulties encountered by the enumerator. Additional information may also have been placed here occasionally by project staff - COMMENTS
Four geographical variables were specified for each record.
CDID: Census District code, comprising initial letter(s) for the province, followed by the sequential three-digit number -- from O001 for Essex in southwestern Ontario to NS206 for Richmond in northeastern Nova Scotia.
CDISTRIC: Census District name, with qualifiers such as NORTH, SOUTH, EAST and WEST placed after the main name.
CSD: Census Sub-District name, with qualifiers such as NORTH, etc were placed after the main name and abbreviated to N, etc. Urban municipal status was also indicated by the addition of T for Town/Ville and V for Village where applicable. In Ontario, township names which duplicated urban municipal names had TP added.
CED: Census Enumerator's Division, a letter and number combination for all or (more often) part of a CSD.
Derived variables were calculated as follows:
TOTEMP: sum of all employees - men, women, boys and girls.
AVWAGE: average monthly wage per employee, calculated only when data for any of the variables WAGES, TOTEMP or MONTH were not missing.
SUMRAWC: sum of all RVALUEs for a given establishment.
SUMPROC: sum of all PVALUEs for a given establishment.
VADD: value added in manufacturing, the difference between SUMPROC and SUMRAWC and calculated only when neither value was missing.
SIC: Standard Industrial Classification code, with up to 11 spaces to allow for variant suffixes and composites of two codes
SEC: major industry group or sector, in which related SIC codes were grouped together.
PROP: "F" denotes a female proprietor.
REFNUM: Unique identification number of record in CANIND71 database.
Procedures in Making Manuscript Information Machine-Readable
We systematized the manuscript details in some cases, to allow for more orderly data retrieval. We corrected obvious spelling errors of common nouns or verbs. We found that errors in transcription of manuscript material could occur with certain letters. In Victorian calligraphy, "S" resembled "F", "M" was hard to distinguish from "N" or "W", and a handwritten double "SS" could resemble a "P", so that "Jane Wissler" looked like "Jane Wipler". We had to be alert for inconsistencies by the enumerators, for example in the order in which they gave surname and forename of proprietors. Errors in numeric values could arise if we misinterpreted the occasional use of a double zero intended by an enumerator to indicate cents as another pair of zeros for the number of dollars. Thus what was intended as $50000 might easily be transcribed and coded as $500.00. Fractions were sometimes used on the manuscript schedules in stating the number of working months in the year and quantities of raw materials and products. We entered these in the computer record as decimals, as in the following examples:
two weeks = 0.5 month
one and a half tons = 1.5 tons
Our procedures in interpreting and transcribing particular variables are explained below:
PROPRIOR
We entered the surname of the proprietor first, followed by the forename. If there was not enough room to enter the names of all partners or the full corporate name, we would add these in the COMMENTS field. Abbreviations were made as follows: Junior(Jr), Senior(Sr), Brother(Bro), Brothers(Bros). If the manuscript census was particularly hard to read, we simply provided the best interpretation of the proprietor's name or at the very least the first initial of the surname or forename followed by a question mark.
In order to achieve the best possible record of industrial activity in 1871, and to overcome some problems of calligraphy in the manuscript schedules, we checked proprietors' names and types of establishment and products for some districts and urban centres. We did this for all urban places in Ontario, for the three census districts of Montreal Centre, Montreal East and Montreal West, for selected rural areas in Ontario, and as spot checks in difficult cases throughout the four provinces. We checked information from the census schedules against lists in the R.G. Dun reference books of credit-worthiness and in city, county and provincial directories and historical atlases. In cases where the census enumerator's spelling of the proprietor's name differed slightly from that in a contemporary printed source, while clearly referring to the same establishment, the record was "corrected" if this would result in a more plausible or conventional rendering of the name. Users may search proprietors' names.
TYPEEST
The census enumerators in 1871 were not provided with any controlled vocabulary for describing types of industrial activity. They entered information about the kinds of industry in the language that came naturally to them or their informants. Thus there is considerable variety in the terms used to describe workplaces and industrial processes.
The establishments in which bread was baked were described in the following ways in English: Bake house, Bake shop, Baker, Baker factory, Baker house, Baker shop, Baker & confectioner, Bakery, Bakery establishment, Bakery shop, Bakery & biscuit factory, Bakery & confectionery, Bakery & crackers, Bakery & pastry, Baking establishment, Baking manufactory, Bread bakery, Bread & biscuit bakery, Bread & cake establishment, Bread & fancy bakery, Confectionery, Confectionery & bakery, and Soft bread bakery.
The activity of making shoes was recorded in the following kinds of establishment in French: Atelier de chaussure, Boutique de chaussure, Fabrique de chaussure, Boutique de chaussures, Manufacture de chaussures, Cordonnerie, Atelier de cordonnerie, Boutique de cordonnerie, Maison de cordonnerie, Cordonnier, Atelier de cordonnier, Boutique de cordonnier, Boutique de cordonniere, Fabrique de souliers, Manufacture de souliers et bottes, Manufacture de souliers.
At first, we considered standardizing all such variants into one term if their product was the same. But, for several reasons, it was decided that the natural language in all its variations should be transcribed with little or no change into the computer record.
1. One factor was our wish to reduce the potential for error in data entry. The task of deciphering and transcribing data from the microfilmed manuscript schedules was demanding enough without asking assistants to make taxonomic judgments at the same time.
2. Secondly, we considered that the natural language, whether in English or French, would have intrinsic interest for some users of the database. Usage of the terms to describe kinds of workplaces, such as "shop", "forge", "manufactory" and "factory" in English and "atelier", "boutique", "fabrique" or "manufacture" in French, was considered possibly significant in contemporary perceptions of industrial settings. Regional variations in industrial language of all kinds were also thought to be interesting. For guides to the natural language used in the 1871 census manuscripts or see the two reports compiled by project staff in the CANIND71 series, Glossary of Industrial Language (#5, 1989), and French-English Dictionary of Industrial Language (#6, 1989).
3. Third, we were also assigning Standard Industrial Classification codes to every record, so there was no need to standardize the terms for data processing purposes.
Thus we transcribed the English words "shop", "forge", "factory", "manufactory" and "works" if there was space available. Such terms for various work environments could be abbreviated if space required. An establishment combining two or more products or services had both keywords entered and linked by a / , and the entry probably had to be abbreviated as well. This abbreviation could take several forms: a pluralization of the second product or service; a shortened form of the words shop, forge, factory, mill, or works; or an abbreviated form of one or both of the main products or services. Common examples included:
Blacksmith/Carriages
Wagons/Blacksmith
Tannery/Boots/Shoes
Saw/Shingle/Grist M (for Mill)
Carriage/Sash/Door F (for Factory)
Brass Fndry/Lamp Fcy
Moulding/Planing Fcy
Blcksmith/Machine Sp (for Shop)
Wagon/Carriage Shop
In cases where the census enumerators added an "s" to certain types of establishments to indicate the possessive, we transcribed without the "s". Thus "Carpenters Shop" became "Carpenter Shop", and "Tailors Shop" became "Tailor Shop". Where enumerators may have used compound nouns to describe the type of establishment, such as "Shoe Maker", "Cabinet Maker" or "Carriage Maker", we contracted these into single words for consistency and brevity. Thus "Cabinet Maker" became "Cabinetmaker" and so on. If the enumerator had entered two establishments with separate numeric data under one proprietor's name, we entered these individually as separate establishments. Often the floating capital was the only joint figure provided and we allocated this between the two establishments in proportion to the value of production of each.
French expressions for kinds of establishment were also slightly modified in data entry. The enumerators in francophone districts usually began with a generic word such as "atelier", "boutique", "manufacture", "moulin", or "fabrique" to describe the type of workplace and followed it with the particular activity. "Atelier" and "boutique" connoted a smaller scale of operation and the use of manual rather than inanimate power, "atelier" being used for the workplaces of artists, dressmakers, florists, cabinetmakers, printers, milliners, pastrycooks, pharmacists, photographers and upholsterers and "boutique" for bakers, hatters, carpenters, wheelwrights, shoemakers, tinsmiths, blacksmiths, goldsmiths, saddlers, tailors, tanners and coopers. "Moulin" was consistently used for the types of powered work settings where "mill" would be used in English, such as a saw mill, flour mill or carding mill, though a woolen textile mill was described as a "Fabrique de drap". The terms "manufacture" and "fabrique" were apparently used for larger scale enterprises, generally equivalent to manufactories and factories in English.
To fit the details in our available space and to give prominence to the particular type of industry, we entered this keyword first and followed it with an abbreviation for the generic term that described the type of workplace. Thus "Atelier de cordonnier" was entered as "Cordonnier, A"; "Boutique de ferblantier" as "Ferblantier, B"; "Moulin à scie" as "Scie, Ml"; "Manufacture de chassis" as "Chassis, Mf"; "Fabrique de meuble" as "Meuble, F"; and "Machine à bardeau" as "Bardeau, Mch". Users may search our list of types of establishment and find individual establishments using the Complex Search Strategies.
TYPEPOW and FORCE We entered the type of power in full and in English -- WATER for EAU, STEAM for VAPEUR and HORSE for CHEVAL -- as we needed to standardize these terms for comparative analysis. If more than one kind of power was given in the manuscript schedules, we abbreviated each to its initial letter. So "Water and Steam" was entered as "W/S". In the case of establishments using horse power, the word "HORSE" might have been entered on the manuscript schedules but the FORCE space left blank. In such cases, we entered the numeral "1" for FORCE. Please use our guide to Complex Search Strategies to search for details of types of power and numbers of horsepower--also possible in combination with other variables.
KINDS, QUANTITIES AND VALUES OF RAW MATERIALS OR PRODUCTS
The 1871 census schedules allowed space for enumerators to complete details of the quantities as well as the dollar values of raw materials and manufactured products. But it was expected that in many cases the raw materials or products would be "lumped together" and given an "aggregate value" or "montant de la valeur". The census enumerators certainly differed in their handling of this part of the schedule, some making considerable efforts to ascertain and record the types, quantities and values of component raw materials and manufactured products. Others simply named one or several materials or products but did not specify separate quantities or values. In the vast majority of records, details for raw materials could be fitted into one or two sets of fields in the computer record, as could details for products. Only for 4.5 per cent of Maritime establishments, 4.9 per cent of those in Quebec and 6.5 per cent of those in Ontario, were three or more sets of fields of the record structure required for products or raw materials.
We took considerable care in transcribing information for raw materials and products. If separate quantities and/or values of raw materials and products were recorded in the manuscript schedules, we entered these separately in the computer record. If only a list of raw materials or products was given and only one aggregate value and quantity, we entered them as a string separated by one or more slashes (/). If separate quantities were given for each raw material or product, but only one value, this aggregate value was entered separately so that it would not be mistaken for a specific value for an individual raw material or product. Sometimes it was not obvious whether a quantity or value referred specifically to one or both of the raw materials or products. In such cases of doubt, the aggregate value only would be entered and not attached to any one specific raw material or product. In some cases, the statement of a general kind of raw material or product was followed by several qualifiers, which we separated with commas, as in these examples:
LOGS,SPRUCE,PINE,BIRCH,HEMLOCK GRAIN,WHEAT,OAT,RYE,OTHER BILLOTS DE PIN,D'EPINETTE,DE PRUCHE,D'ERABLE GRAIN,BLE,SEIGLE,AVOINE,ORGE,SARRASIN Lengthy phrases such as "of different kinds" or "of various kinds" were usually replaced by "ASSORTED" or "ASSORTIS".
Users may search our lists of products and raw materials and use our guide to Complex Search Strategies to find enterprises reporting specific materials or products, perhaps in combination with other search variables.
UNITS OF MEASUREMENT
We entered units of measurement for raw materials and products only if they were actually stated. We also transcribed units in the language used by the original enumerators. No attempt was made to standardize equivalent measures which were given different names in English and French, such as "hundredweight" or "cwt" in English and "quintaux" in French. Consistent abbreviations were used for units of measurement where they were longer than six characters. See also: English and/or French Units and Abbreviations.
COMMENTS
Information appearing on the Remarks/Remarques column of the schedule was entered in the COMMENTS field, together with additional information from any other column. Two hundred character spaces were allowed in the record structure for this purpose. A few zealous enumerators entered more precise locational information in the Remarks column, such as street address in some cities or locality name in a few rural districts. General comments relating to the whole census enumerator's division were sometimes written on the last of the manuscript schedules for that particular division. Altogether, 14 per cent of the establishments in New Brunswick and Nova Scotia have some details in the COMMENTS fields, and 21 per cent of those in Ontario and Quebec.
Our team also used the COMMENTS field to add information that could not be fitted in the proper fields. Most often this added details of partners or corporate names to what could be fitted in the PROPRIOR field or amplified the description of the type of establishment. In a very few cases, additional raw materials or products were specified, if there had been inadequate space for them in their proper fields.
In some sample districts, we also noted in the COMMENTS field whether information in any field had been apparently changed at all from what was originally entered by the enumerator. Such changes seem often to have been made in another hand, perhaps by the District Commissioner or the central office staff in Ottawa. Our practice was to enter the data as changed or corrected, but we were interested in assessing the extent of such changes. The sample districts were Ontario Census Districts 54 to 58 and 63 to 66 -- comprising the counties of Northumberland, Peterborough, Lennox, Addington and Frontenac and the City of Kingston -- and also the Quebec Census Districts 109 (Vaudreuil), 110 (Soulanges), 143 (Portneuf) and 157 (Beauce Ouest). For all records in these districts, we added amendment codes in the COMMENTS field if data in any other field had been amended. The following system of abbreviations was used:
A at the beginning of the code indicated an amendment of some kind
E indicated some change in data for employees
FIC indicated some change in data for fixed capital
FLC indicated some change in data for floating capital
M indicated some change in data for months
P indicated some change in data for products
Q indicated a change in the quantities of raw materials or of products
RM indicated some change in data for raw materials
V indicated a change in the values of raw materials or of products
The commonly used code of ARM-Q thus meant that the details for quantity or quantities of raw materials originally entered by the enumerator had been changed. We found that 21 per cent of all the records in the Ontario sample districts had some change in their data and 26 per cent of the records in the Quebec sample districts.
Back to topCANIND71: alphabetical list of variables
Alphabetical list of variables.
Back to top
GLOSSARY OF INDUSTRIAL LANGUAGE
Glossary of Industrial Language.
Units of Measurement and abbreviations
See also:
English Units and Abbreviations
French Units and Abbreviations
Standard Industrial Classification Applied to 1871 Census
The Standard Industrial Classification of 1970 was used to categorize the industrial establishments of 1871, and SIC and SEC codes were added to each record.
Until the Dominion Bureau of Statistics was founded in 1918, Canadian census takers made virtually no attempt to classify industry types in any sort of logical system. The determination of "kind of establishment" was left to the responding business and the enumerator who were not bound by any prescribed vocabulary or terminology. The tabulating clerks in Ottawa grouped similar types as best they could. Summary details of industry types were published in five series in 1871, each in alphabetical order of the English name. As the census reported, the series distinction had "no apparent importance except that of convenience". It seems to have been more convenient for typographical layout to group industry types found in all, some, or a few Census Districts. Series 1, for example, grouped industry types that were found in practically all Census Districts. At the other extreme, Series 5 grouped industry types found in fewer than 20 Census Districts and sometimes in only one. These alphabetical listings (based on the English names of the industry types) pose problems for users who wish to make comparisons with other places and periods, as there is no indication of what precisely is included or excluded. In an industry such as agricultural implements, for example, it is not clear whether hand tools are included with the machines.
For a detailed discussion of standard industrial classifications and of the relationships among Canadian, American and British practices, see Standard Industrial Classifications Applied to Historical Data: The Case of the 1871 Industrial Census, #7 in the CANIND71 series of research reports (1989).
From the beginnings of the CANIND71 project in 1982, we realized that we should supplement the natural language used in the manuscript schedules with systematic coding of industry types. The 1970 Standard Industrial Classification Manual was the obvious choice for adoption. We had used the manual for many earlier projects ever since it had first appeared in December 1970 so that we were familiar with the structure and nuances of the classification. The 1980 SIC system did not come into use in the Statistics Canada Annual Census of Manufactures until 1984 for the 1983 reference year and it was only about this time that the first 1981 census reports using the 1980 SIC classes began to appear.
Our adoption of the 1970 system at the beginning of the project provided an excellent classification system that was both comprehensive and thoroughly tested. The 1970 Manual was well organized with extensive definitions of each industrial group. In addition, the volume had very detailed numerical and alphabetical indexes, so that there were few difficulties in classifying any establishment to a particular industry Group. The CANIND71 project used 86 of the Groups in all but one of the Major Groups. Only one new number (242) had to be used to cover a group not differentiated in the 1970 SIC. In all, 27 of the 1970 SIC groups were not used in CANIND71 since these industries did not exist in the 1870s.
Understandably, there were some problems in applying a 1970 classification to the conditions of a century earlier.
1. First, changes in technology have eliminated some industry types that were common in 1871. We compiled a glossary of industrial language as a guide to the precise classification of obsolete industrial activities, processes and products. Equivalent French terms are listed are listed in French-English Dictionary of Industrial Language, #6.
2. Second, the more rudimentary level of economic and business organization in 1871 meant that many enterprises combined industrial activities that would more typically be separated in specialized businesses by 1970. Third, the repairing of all kinds of goods and equipment was more integrally associated with manufacture in 1871.
Each of the 45,070 records in the CANIND71 database was given both an SIC code for its specific industry type and an SEC code for the major industry group into which specific types may be generalized (Table 6.2). The addition of the standardized industry codes to records in the database is obviously necessary for the retrieval and analysis of firms in the same or related lines of business. There were other advantages. With a standard industrial code for each record, we did not need to try to standardize the verbal details for type of establishment that had been entered by the original enumerators. The SIC classification also enables the user of the CANIND71 database to cope with the full variety of establishments enumerated in the 1871 census, when a wider range of activities and processes was included in definitions of industry than would be today. As well as the major group for manufacturing itself (SEC 5), these include agricultural services (SEC 1), forestry (SEC 2), fishing (SEC 3), mines, quarries, oil and salt wells (SEC 4), construction (SEC 6), utilities (SEC 7), trade (SEC 8), and services (SEC 10). By excluding establishments with these SEC codes, it is possible to create more select datasets from the original database for comparative purposes, limiting one's attention to activities that are now more strictly defined as manufacturing industry.
We adapted the 1970 system to the 1871 context in two main ways. First, suffixes of hyphen and capital letter might be added to the basic 3 digit code to give greater specificity. A brewery (109 B) was distinguished from a distillery (109 D) for example. Second, an establishment that combined two or more products or services was given a compound SIC code consisting of the two most important linked by a slash (/), so that a flour mill cum distillery is represented as 105/109 D. These elaborations of the basic system resulted in a total list of over 720 specific codes that were used to describe all the 45,000 firms in the four provinces in 1871.
Common compound codes that were used at least 40 times in the CANIND71 database to describe enterprises that combined two or more principal types of products were:
105/105-O: flour and oatmeal mill/moulin à farine et farine d'avoine
107/108-C: bakery and confectionery/boulangerie et confiserie
179-S/179-T: saddlery and trunkmaking/sellerie et fabrique de valises et boîtes
244/249-M: dressmaking and millinery/habillements pour femmes et modiste
249-M/244: millinery and dressmaking/modiste et habillements pour femmes
251-S/259-C: shinglemaking and cooperage/confection de bardeaux et tonnelerie
251/251-S: moulin à scie et confection de bardeaux
254/261: sash,door and blind factory combined with furniture and cabinet making/fabrique de portes et fenêtres et meublerie
261/258: cabinet and coffin making/meublerie et cercueils
261/329: cabinet and carriage making/meublerie et carrosserie
307-S/311: manufacture of stoves and agricultural implements/fabriques de poêles et d'instruments aratoires
311/329: agricultural implement and carriage making/fabrique d'instruments aratoires et carrosserie
329/896: carriage making and blacksmithing/carrosserie et forges
The appropriate SIC code was determined by study of both the stated kind of establishment and the actual products and raw materials reported in each case. Wherever possible, we chose the SIC code that best matched the actual outputs of the business as reported for the year ending March 31, 1871. The "kind of establishment" information alone could be quite misleading, as this term was sometimes an occupational label which did not really describe the actual industrial processes and products. Thus a woman might be described as a "milliner" but if the products of her business included only dresses and other women's clothing rather than bonnets, hats and women's head-dresses, she would be assigned a code of 244 for general women's clothing and not 249-M for millinery. Or a business might be described very generally as a "foundry" which would rate a code of 294 only if its products were primarily iron castings. But more typically its products would be fabricated metal products such as stoves (307-S) or general machinery (315) or agricultural implements (311). Or one who was by trade a blacksmith and so recorded in the "kind of establishment" may have had opportunities to make all kinds of metal goods, often including carriages and wagons. If such a person made definite numbers of carriages or other items of transportation equipment, his business would be assigned the appropriate and more specific SIC code of 329 rather than 896 which was left for a blacksmith engaged only in "country work and repairs" or "ouvrages assortis".
Click here for codes for the major industry groups (SECs) with the range of specific industry types (SICs) included in each.
We also present tables of the basic SIC codes that were assigned in making the 1871 industrial census data machine-readable (including suffixes but not compunds).
They are presented in three systems of order:
a) By logical order of the SIC itself. In this table, the terms that were also used in the published 1871 census reports are printed in italics.
b) By English names.
c) By French names.
Back to top
Suggested Citation of CANIND71
The source of all data, documentation or programs derived from the
CANIND71 database should be acknowledged as:
Canadian Industry in 1871 Project (CANIND71). University of Guelph, Ontario, 1982 - 2008. After the first reference to the full citation in each work by a user, the short form "CANIND71" may be used for subsequent references.
Back to top
