STATE AND LOCAL STATISTICAL STUDIES CONDUCTED
AS WORK RELIEF PROJECTS*
FREDERICK F. STEPHAN
Federal Emergency Relief Administration
ARIZONA is completing an unem¬
ployment survey and school cen¬
sus, California is studying the
coses of medical care, Washington has sur¬
veyed its state welfare system, Colorado is
* This paper was read before a joint session of the
Section on Social Statistics of the American Socio¬
logical Society and the American Statistical Associ¬
ation at the annual meeting in Chicago, December,
1934-
studying unemployment insurance, North
Dakota is analyzing local taxation and
finance, Missouri is appraising its school
facilities, Illinois is studying transients
and population movements, Alabama is
examining the occupational characteris¬
tics of its relief population, Pennsylvania
is making a child welfare survey, Mary¬
land has conducted a crime prevention
study, Florida measured the attitudes of
486
SOCIAL FORCES
relief clients, and New Hampshire is sur¬
veying the cost of living. Hundreds of
projects covering a wide range of subjects
in the field of social statistics are being
conducted throughout the United States as
part of the work relief program of the
Federal Emergency Relief Administration.
The state and local emergency relief
agencies are responsible for the actual ad¬
ministration of both direct relief and work
relief; the Federal Emergency Relief Ad¬
ministration supplies part of the funds,
sets up standards and policies, and per¬
forms numerous advisory, planning, and
coordinating functions. In the case of
statistical and survey projects, each appli¬
cation is reviewed in its statistical aspects
by the Coordinator of Statistical Projects
who furnishes advice on technical prob¬
lems, suggests improvements in procedure,
and coordinates projects to prevent over¬
lapping, duplication and conflict. The
coordination function is a very significant
activity but this paper will be limited to
a discussion of the projects themselves.
The background for these projects may
be traced to the first two or three years of
the depression when private emergency as¬
sociations and municipal relief organiza¬
tions carried the load of unemployment
relief. In New York, Philadelphia, Chi¬
cago, and many other cities, ' 'made work''
was provided as a form of relief preferable
to a simple dole or direct relief. In con¬
nection with these work programs a num¬
ber of statistical and research studies were
undertaken. As the depression deepened,
the States and later the Federal Govern¬
ment assumed part of the responsibility
for aid to the unemployed and set up work
relief programs.
The climax of this movement to date
was the organization of the Civil Works
Administration which provided work for
4,000,000 unemployed workers last winter.
Many statistical projects were undertaken,
including population and unemployment
censuses, price studies, housing inven¬
tories, health surveys, traffic studies, a
national census of business, studies of tax
delinquency, and relief surveys. More
than $10,000,000 was expended for special
statistical studies conducted by the Federal
departments and agencies alone. The
CWA statistical studies were conducted
very much as are most surveys not con¬
nected with relief and unemployment.
With the exception of the haste with
which they were organized, they had rela¬
tively few of the restrictions which have
surrounded work relief projects.
Like all work relief projects, the statis¬
tical and survey projects have been organ¬
ized principally to provide useful work for
unemployed white-collared workers who
are eligible for relief; their contribution to
social statistics is a by-product, that is, it
is secondary to the purpose of aiding needy
unemployed workers. In this respect they
differ from the research projects described
by the foregoing speakers. The statistical
projects are almost exclusively planned
and initiated by state, county, or munici¬
pal governmental agencies, frequently in
association with private organizations.
General supervision, office space, materials
and equipment are usually furnished by
these project sponsors; the local relief ad¬
ministration furnishes the workers drawn
from its relief families or from lists of
clerical, professional, and other white-
collared workers who are certified as eligi¬
ble for relief but who have not applied for
it. With few exceptions the workers are
restricted to a number of hours per week
which will earn the minimum budgetary re¬
quirements of their families. The proj¬
ects are similar in organization to other
work projects of a non-statistical character
and while they are not "made work,"
their relief features introduce a number of
limitations which could be avoided if their
STATISTICAL STUDIES AS PROJECTS
487
principal purpose was to produce statisti¬
cal information.
The restrictions to which work projects
are subject are dictated by the fact that
work relief is relief and not reemployment,
a substitute for direct relief and not for
unemployment, principally a means of
preventing suffering and need and second¬
arily a means of producing a valuable prod¬
uct. For this reason one is justified in
regarding the statistical results as a by¬
product and in counting them very largely
as net gain since a very large portion of
their total cost would have been spent in
direct relief if they had not been under¬
taken.
It is interesting to note that many im¬
portant statistics are by-products of non-
statistical activities. Mortality statistics
are a by-product of the regulation of buri¬
als; income statistics, of the income tax;
and the population census, of representa¬
tive government. It is conceivable that
the development of these projects and the
recognition of their potential values might
lead to a series of continuing projects in
social statistics maintained as part of a
program of work relief or public works,
unless all the competent clerical and pro¬
fessional workers are absorbed by regular
employment.
TYPES OF STATISTICAL PROJECTS
The projects which have been reviewed
by the Coordinator of Statistical Projects
cover a wide field of social and economic
statistics. A tabulation of the first 743
projects reviewed favorably revealed the
following distribution:
Number
of projects
Type of project reviewed
Comprehensive Planning and Social
Surveys 56
Governmental Organization and Taxa¬
tion 39
Education and Schools 63
Social Welfare and Relief 106
Number
of projects
Type of project reviewed
Health and Sanitation 59
Population and Vital Statistics 63
Occupations, Employment, Unemploy¬
ment 40
Agriculture 51
Prices, Business and Industry 34
Housing, Mortgages, and Land Utili¬
zation 115
Traffic 74
Historical Research and Records 41
74J
Within each of the above categories may
be found wide variation in the scope,
methods, and emphases of the projects in¬
cluded. "Social Welfare and Relief" in¬
cludes a state-wide survey of welfare agen¬
cies and facilities in Colorado, a study of
unmarried mothers in Minnesota, a com¬
pilation of the statistics of charity granted
by the Church of Latter Day Saints in
Utah, and an analysis of rural relief cases
in certain Pennsylvania counties. ' 'Popu¬
lation" includes a study of residential mo¬
bility in Indianapolis, a tabulation of
births and deaths by census tracts in Louis¬
ville, preparation of an index of births,
deaths and marriages in Newport, Rhode
Island, and a census of the blind in Brook¬
lyn. The other classes of projects reveal
a corresponding variability.
A number of these projects are suffi¬
ciently interesting or unique from the
standpoint of social statistics to merit de¬
scription in detail but only a few can be
selected in a paper of this length.
PHILADELPHIA HOUSING SURVEY
Philadelphia was not one of the 64 cities
included in the Real Property Inventory
Project taken under CWA by the U. S.
Bureau of the Census and Bureau of For¬
eign and Domestic Commerce but it con¬
ducted a housing survey which was similar
to the Real Property Inventory being more
488
SOCIAL FORCES
detailed on the general physical character¬
istics and less detailed on financial items.
The survey was started as a CWA project
with more than 1700 workers in the field
and is now being completed as a work re¬
lief project employing 150 tabulators and
clerical workers. It included data on
450,000 structures covering the type of
buildings, age, condition, use, ownership,
sanitary and heating facilities, occupancy
of each household unit, value and rental,
concessions and services, previous resi¬
dence of occupants, and many other items
of social and economic significance. Ma¬
chine tabulation of the schedules makes
feasible detailed cross tabulations includ¬
ing some counts by blocks and by census
tracts. More intensive investigations of
slum areas are closely correlated with the
city wide survey.
TACOMA SOCIO-ECONOMIC SURVEY
Social, medical, and economic data for
17,000 households were secured by an
elaborate inquiry conducted by the Wash¬
ington Emergency Relief Administration,
the Tacoma (Washington) City Planning
Commission, the Tacoma City Health De¬
partment, and the Foundation for Social
Research in Medical Care. Information
was secured for each member of each
household covering the usual census items,
religion, income, recreation, illnesses and
medical care. Surprisingly little difficulty
was reported, due to careful preparation,
highly selected workers and other favor¬
able factors. The data are to be analyzed
over a period of two years. Private funds
have been allocated for this purpose and
for a share of all other costs.
Both the Tacoma and the Philadelphia
projects illustrate the great detail with
which social statistics have been collected
in some of the more exceptional work
projects. Many other projects have ap¬
proached them in the range of items which
were secured and tabulated. Projects
sponsored by State Planning Boards and
State Welfare Surveys in several states
have covered broad fields of social data
but have relied principally on information
already collected.
DAYTON OCCUPATIONAL CHARACTERISTICS
SURVEY
An unemployment census project sub¬
mitted by Dayton, Ohio, offered an oppor¬
tunity to set up a cooperative relationship
between it and the occupational charac¬
teristics survey mentioned by the preced¬
ing speakers. The schedule was adapted
by suitable additions to meet the needs of a
local census and a relief-nonrelief compari¬
son. The latter phase of the survey pro¬
vided data on nonrelief workers and their
families comparable to the data secured for
workers on relief. The editing and tabu¬
lating of the schedules was made uniform
with that for the other cities in the Survey
of the Occupational Characteristics of Per¬
sons Receiving Relief and thereby Dayton
and the FERA secured important com¬
parative statistics.
PORTLAND SCHOOL CENSUS AND UNEMPLOY¬
MENT SURVEY
Another instance of cooperation in an
unemployment census is the Portland, Ore¬
gon, Project in which the regular school
census was merged with a special unem¬
ployment survey needed by the Unemploy¬
ment Insurance Commission. The funds
usually expended for a school census were
contributed to the larger survey and the
regular school census data were taken
from the schedules.
NEW HAMPSHIRE FAMILY EXPENDITURES
STUDY
A further study of family expenditures
of wage earners sponsored by the New
Hampshire Minimum Wage Office and the
STATISTICAL STUDIES AS PROJECTS
489
United States Bureau of Labor Statistics.
The study is tied in with a larger study to
provide revised weights for the BLS cost
of living index. About 1150 detailed fam¬
ily expenditure schedules are being se¬
cured from various parts of the state.
Exceptionally favorable conditions of
planning and supervision make feasible an
inquiry which lists almost xoo food items,
150 clothing items and correspondingly
detailed information on other fields of
family expenditure.
LOUISVILLE CENSUS TRACT TABULATION
The Louisville Department of Health
secured five clerical workers to draw up
census tract areas and to tabulate births,
deaths, and reported communicable dis¬
eases by census tracts for use by the De¬
partment and by other social or health
agencies in the city. Preparatory work
over a period of three years was utilized in
dividing the city into census tracts and the
project was conducted within a continuing
statistical agency.
MICHIGAN UNEMPLOYMENT SURVEY
Plans are now being perfected for a sur¬
vey of unemployment in Michigan which
is notable for the sampling methods which
are proposed.
In parts of Detroit one tenth of all ad¬
dresses are to be taken according to pre-
designated house numbers. In difficult
areas a complete enumeration will be
made. Similar methods will be used in
other cities and for the rural areas sample
townships will be completely enumerated.
Detailed estimates will be computed for
the state, for cities, and for the rural part
of the state.
This Michigan sampling procedure and
schedules developed in Ohio, Arizona, and
elsewhere are instances of experiment in
statistical methods which reflect the op¬
portunities opened by work relief projects
for the development of social statistics.
The preceding thumb nail sketches
merely suggest the character of a few of the
more noteworthy projects. Without in¬
dicating the supporting evidence, I will
list some of the difficulties encountered by
statistical work projects:
1. No mobilization plans have been
available to guide the rapid development
of statistical projects. University pro¬
fessors and city department heads, who
had had only meager experience in organ¬
izing and supervising large-scale surveys,
were suddenly made field marshals in com¬
mand of fifty or a hundred enumerators.
In many instances hastily drawn defini¬
tions, instructions, and schedules proved
to be ambiguous or inadequate, the time
needed to complete projects was under¬
estimated, the necessary supplies and fa¬
cilities were not all anticipated, and
perhaps most frequently there was no
preparation for summarizing or tabulating
the data after they were secured. These
weaknesses were such as one might expect
under the circumstances and there was
usually an earnest effort to correct them
once they were discovered.
x. A number of projects have had ex¬
pert supervision but a larger number were
directed by persons who had had little
direct contact with the collection and
analysis of social statistics. It would
have been unreasonable to expect to find a
large enough number of competent statis¬
ticians to meet the requirements of an
emergency program. If statistical activi¬
ties continue to expand, the next emer¬
gency period will probably find a more
nearly adequate supply.
3. Another weakness has been the gen¬
eral shortage of more capable workers due
to their greater reluctance to apply for re¬
lief and also the short hours they worked
SOCIAL FORCES
496
to cam their budgetary allowance. In ad¬
dition, there was almost constant uncer¬
tainty as to how long it would be before
some man in a key position would leave
for a regular job or the whole project be
suspended or terminated by changes in the
relief situation.
4. One of the principal weaknesses of
the statistical projects (and one they share
with many surveys undertaken by public
and private agencies with no relation to
relief) has been the difficulty of dissemi¬
nating the findings and of putting them to
use. Funds for publication have seldom
been available. The staff of workers usu¬
ally disbands as soon as the project is com¬
pleted. There is often no continuing
organization equipped to follow up the
survey. In most cities there are great
opportunities to preserve and assemble the
results of several surveys and to exploit
them both for further research and for the
public welfare. Municipal research and
planning agencies and the universities are
expressing interest in these possibilities.
Numerous and varied contributions are
being made by work relief projects to
social statistics. Great masses of data of
varying quality are being assembled and
made available for analysis and interpre¬
tation. People are becoming accustomed
to social and economic surveying. Ad¬
ministrators and officials are seeing oppor¬
tunities to use statistical information in
planning and routine administrative prac¬
tice. Experiments are being made in the
procedure of securing and analyzing statis¬
tical data. Finally, these projects tend
to maintain the morale and welfare of the
group from which will be drawn many of
the workers needed by any great future
expansion of social statistics in other than
emergency agencies.
■ # p IJvHPSS
/ v -