With Respondents and Would Increase Costs on an Already Costly Survey, But

DOCYMFNT RF l'MF ED 022 025 VT 005 212 MANPOWER FORECASTING THROUGH THE OCCUPATIONAL NEEDS SURVEY. Rutgers, The State Univ., New Brunswick, N.J. Inst. of Management and Labor Relations. Note- 45p EDRS Price MF-$0.25 HC-$1.88 Descriptors-*EMPLOYMENT PROJECTIONS, *EVALUATION, MANPOWER NEEDS, *OCCUPATIONALSURVEYS, RESEARCH METHODOLOGY Identifiers-New Jersey To test the efficacy and validity of measuring future manpower needs by means of an employer survey, employer predictions of employment needs as reported inthe 1960 Area Skills Surveys, done by the New Jersey Division of EmploymentSecurity, were compared with actual employmentreported in the Occupational Training Needs Survey conducted in the Newark Labor Market Area in late 1963. The1963 survey provided data from 337 of 604 employers contacted on September 1963 employment and estimated expansion needs as of September 1965 and September1968. On a followup form in the fall of 1965 showing their estimated September 1965 employment, 302 respondents indicated actual September 1965 employment, and checked reasons for any discrepancy between the two. Data suggested that, in overall terms, estimates of changes in total employment were reasonably accurate. It was concludedthat the estimates collected in 1963 might have served with other availableinformation provide as solid a basis for occupational manpowerneeds as might be wished, the imperfect estimate was better than none. Improvements would require morepersonal contact with respondents and would increase costs on an already costly survey,but the enhanced accuracy might make the effort worthwhile. (MM) RESEARCH SECTION

INSTITUTE OF MANAGEMENT AND

LABOR RELATIONS

MANPOWER FORECASTING through the occupational needssurvey

RESEARCH SECTION INSTITUTE OF MANAGEMENT AND LABOR RELATIONS UNIVERSITY EXTENSION DIVISION RUTGERS - THE STATE UNIVERSITY U.S. DEPARTMENT OF HEALTH, EDUCATION & WELFARE

OFFICE OF EDUCATION

THIS DOCUMENT HAS BEEN REPRODUCED EXACTLY AS RECEIVED FROM THE RESEARCH LIBRARY

PERSON OR ORGANIZATION ORIGINATING IT.POINTS OF VIEW OR OPINIONS THE CENTER FOR VOCATIONAL

STATED DO NOT NECESSARILY REPRESENT OFFICIAL OFFICE OF EDUCATION AND TECHNICAL EDUCATION -.111E-OHMSTATE-UtlIVERSITY POSITION OR POLICY.

RESEARCH SECTION

INSTITUTE OF MANAGEMENT AND

LABOR RELATIONS

_MANPOWER FORECASTING through the occupational needssurvey

RESEARCH SECTION

-3INSTREITE-OF MANAGEMENT AND LABOR RELATIONS UNIVERSITY EXTENSION DIVISION wt.-. .%). 1 RUTGERS2THE STATE UNIV-ERStrt Jas, SEPTEMBER, 1966 Acknowledgements

This study would not have beenfeasible without the cooperation of the NewJersey Division of EmploymentSecurity, and particularly that ofMr. Walter Chartier and Mr.Al Davis of the Division's Bureauof Statistics and Research. Mrs.

Estelle Blostein assisted inthe processing and tabulationof data. However, the interpretations andconclusions are solely the responsibility of theundersigned.

Jack Chernick, Chairman Research Section Institute of Management and Labor Relations

September, 1966 TABLE OF CONTENTS Page

I Introduction 1

II Findings 8 Predicted Expansion Needs 13 The Quality of Employer Response 17 Reasons for Inaccurate Forecast 23

IIISummary and Conclusions 26 Appendix 31 A - Questionnaires 31 B - Employment By Occupation and Industry for September, 1963 41

LIST OF TABLES Number Page

1. Skill Surveys Conducted In New Jersey, 1960 to 1963 2 2.Summary of Company Responses To Follow-Up Study 5

3. Employment Changes By Industry, Newark Labor Area and Sample Survey Compared September, 1963 to September, 1965 7

4.Actual. Employment Compared With Predicted Employment for September, 1965 And With Employment In September, 1963 By Industry Group 302 Establishments In NewarkLabor Area 9

5.Employment In September, 1963 And Predicted And Actual Employment In September, 1965 By Major Occupational Classifications 10

6. Predicted And Actual Employment Compared By Major Occupation And Industry Groups 12

7.Predicted And Actual Changes In Number of Employees Between September, 1963 And September, 1965 For Selected Industry Groups, By Major Occupational Classifications 15

8. Employment Changes In Selected Occupations September-1 1963 to September, 1965 18

9.Number And Median Size Of Employment In Establishments By Industry, Classified According To Whether Or Not Predictions Were Made 21

10.Reasons For The Difference Between Expected AndActual Employment As Of September, 1965 And Number Of Employees Affected 24

A.Employment By Occupation And Industry For September, 1963 Predicted And Actual, September, 1965 42

11; _

MANPOWER FORECASTING THROUGH

THE OCCUPATIONAL NEEDS SURVEY

INTRODUCTION

Beginning in July, 1960 the N.J. Divisionof Employment

Security launched a series of AreaSkills Surveys designed to disclose expected manpower needs ofemployers in the various labor market areas of New Jersey.

The primary purpose of the surveys, asexplained lo employers, was to determine existing andanticipated job oppor- tunities and to compare these with theexpected output of existing programs of training. The ultimate purpose, of course, was to achieve a greaterequalization between demand and supply of manpower in the labor market.

In many ways, local EmploymentService Offices can function efficiently only if future occupational needs canbe anticipated.

Counselling workers and advising employersinterested in expand- ing in-plant training programs presupposethat Employment Ser- vice personnel know what is likely to occurin a particular labor market area. The Area Skills Survey is one devicefor accomplish- ing such forecasting. A list of the completed surveys isshown in Table 1. Based on information contained in theseveral survey

reports, the table also indicatesthe extent of coverage of estab-

lishments. SKILL SURVEYS CONDUCTED IN NEW JERSEY, Number of TABLE 1 1960 TO 1963 Number of Newark Area Sept., 1963DateSurvey of Questionnaires Mailed1,466 Returned Number 811 ReturnedPercent 55 Employees137,723 Covered Total Employment* Percent of 23 PerthTrentonBridgeton,Paterson Amboy, Vineland New March,March,July, 1963 1962 1960 --- 658274 246392 606673 31,475 ,MA 3736-- Brunswick Sept., 1961 800 517 60 76,526 42 ,MA , 1961 249 64.5 ,MA Atlantic City Nov., 386 - Lakewood,Long Branch Toms River March,July, 1962 1962 286656 452193 67.569 OM OM MO Ocean City, Cape May,Wildwood March,July, 1962 800185 498133 6272 76,500 OM *TheCamdenJersey total City does not include employment in governmer.-., May, 19611962 1,432 884 railroads, mining, forestry, and fisheries. 61.7 --- 40 Source: covered.OnlyReports in theof thesurveys New Jerseyshown wereDivision data ofpresented Employment with Security respect to describing each of the several surveys. the number and percent of employees How accurate are suchforecasts? The present study at- tempts to provide apartial answer to this question bycomparing employer predictions with actualemployment. Selected for the test was the OccupationalTraining Needs Survey conducted in the

Newark Labor Market Area in late1963.* At that time a form** was sent to a sample of establishmentswith the request that employers record employment as of September,1963 and estimate expansion needs as of September, 1965, andSeptembero 1968. Respondents were also asked toindicate the number of replacements expected to occur within the two timeperiods and to indicate how many workers were at that time involved inin-plant training programs.

They were asked to supply this informationfor the establishment as a whole and also for somefifty occupations. The occupations included were those generally in short sup?lyand which reflected the area's occupational distribution. They were selected in ad- vance following reviewof labor market reports of the Divisionof

Employment Security and the local EmploymentService offices.

In the present study the accuracy ofthe estimates supplied

by employers was tested in respect topredicted employment as of

September, 1965. In October of that year employers whohad re-

*For a report on the survey see: Division of Employment Security, Department of Labor.and Industry, State of New Jersey. Occupational Training Needs Survey: Newark Labor MarketArea, 1964.

**A copy of the form appears inAppendix A sponded to the original questionnaire were sent afollow-up

form,* which reproduced their estimates of anticipated employment in September, 1965. Respondents were asked to indicate their actual employment in September,1965 and to explain any differences between estimated and actualemployment

by checking one or more items in a list of possible reasons.

The survey conducted in 1963, consisting of1,466 ques-

tionnaires mailed by the Division of EmploymentSecurity,

yielded 811 returns. In the present study all establishments

with more than ten employees and one-tenth of thosewith ten

or fewer employees receivedfollow-up forms. Thus a total of

604 questionnaires were mailed between October 15 and December

31, 1965. In the case of nonrespondents follow-up letters were

sent. In addition, by means of telephone calls employers were

urged to complete and submit the forms.

A summary of the response rate is indicated inTable 2.

It shows that some 50 percent of all the questionnaires sent

out were returned. In terms of size of employment as of Sep-

tember, 1963, the follow-up survey accounted for some 77 percent

of the work-force covered in the original survey. However, it

should be noted that the original survey, which yielded datafor

approximately 138,000 workers in the area, covered only 23 percent

*A copy of the form appears in AppendixA -5-

exclusive of govern- of the area's wageand salary employment, ment, railroads,mining, forestry, andfisheries.

TABLE 2

SUMMARY OF COMPANYRESPONSES TO FOLLOW-UPSTUDY

Number of Establishments

TenEmployees More than Total Original Survey (Sept.1963) orless Ten Employees 811 Questionnaires returned 240 571 137,723 Employees covered

Follow-up Survey (Sept.1965)

568 604 Questionnaires mailed 36 337 Questionnaires returned 19 318 302 Usable questionnaires(total) 15 287 102,025 Employees covered 50.0% Percent of numbermailed 42% 51% 76.7% Percent of employeescovered in original survey

shows con- The distribution ofreturns in the follow-up survey

siderable variation inthe success with whichthe several industry changes in employmentby groups weresampled. Table 3 compares the in the Area as awhole. industry group in thesample returns with those

sample statistics divergesub- It will be notedthat in some cases the This is true stantially from thefigures for the Area as awhole.

and retail. In the con- of constructionand trade, both wholesale showed a rate of tract constructionindustry, sample establishments that for the increase in employmentapproximately twice as great as actual industry as a whole;in retail trade, oursample suggested an -6-

decline in employment while the totalfor the industry increased by 6.5 percent in the period September,1963 to September, 1965.

These results are partly explainedby the low sampling ratios in these industries. The final column in Table 3 indicatesthat as of September, 1965, only 3.4 percentof employees were covered by sample returns in construction andwholesale trade, and only

6.8 percent in retail trade. In contrast, sampling ratios in the other industry groups ranged from14.6 percent in the service industries to 41.2 percent in transportation,communication, and public utilities. EMPLOYMENT CHANGES BY INDUSTRY, NEWARK SEPTEMBER, 1963 TO SEPTEMBER, (numbers in thousands) TABLE 3 LABOR AREA AND SAMPLE SURVEY 1965 COMPARED NEWARK LABOR AREA SAMPLE SURVEY PercentSample inof Sept., 1963* Sept., 1965* Number Increase Percent Sept., 1963 Sept., 1965 Number Percent Sept., 1965 Increase Universe ' 4 Manufacturing 240.2 (1) 248.6 (2) -UT- 8.45.1 16.8 3.5 (4) 43.1 (5) .9 44.3 1.2(6) -177- 1.2 .3 -757-33.3 2.8 17.8 3.4 (9) Transportation,ContractRetailWholesale TradeConstruction Trade Communi-cations and Public Utilities 42.130.351.990.7 44.052.935.496.6 2.95.91.0 6.56.91.9 21.7 6.81.5 21.8 6.61.5 -.2**-2.9**.1 o 0.00.5 41.2 3.46.8 Finance,Service Insuranceand Miscellaneous andReal Estate TOTAL 607.5104.8 47.5 640.4ii3.7 49.2 32.9 8.91.7 8.53.65.4 102.1 15.712.4 104.7 16.612.7 2.6 .9.3 2.42.55.7 25.816.314.6 *Nonagricultural**DecreaseBureau Payroll of Statistics Employment. and Records, Trenton, New Jersey. Newark Area 1958-1965. March, 1966. New Jersey Department of Labor and Industry, I I

FINDINGS

Clearly, if the goal of the Occupational Needs Survey was to estimate overall change in employment in the Newark Labor Area, the survey conductell in September, 1963 was highly successful.

Thedata reproduced in Table 4 show that actual employment in

September, 1965 in the 302 establishments constituting the sample was almost precisely what had been predicted in September,1963.

Employment in the two-year period increased by three percent; and allowing for some rounding of the figures, the change coincided with predictions made in September, 1963.

However, while of some value, the accuracy of the overall estimate is only of secondary importance. How useful the estimates are as guides in the formulation of manpower policies, in counselling and training programs, depends on the quality of prediction of future needs in given occupations. In this respect, too, the forecasting effort under review may be said to have met with some success: it achieved a reasonable level of accuracy in estimating employment by major occupational group in the 302 establishments. The last column in Table 5 shows actual employment in September, 1965 in percent of predicted employment for that date. The largest discrepancy occurs in the estimate for workers TABLE 4

ACTUAL EMPLOYMENT COMPARED WITH PREDICTED EMPLOYMENT FOR SEPTEMBER,1965 AND WITH EMPLOYMENT INSEPTEMBER, 1963 BY INDUSTRY GROUP 302 ESTABLISHMENTS IN NEWARK LABORAREA

Employment Employment Sept., 1965 in Sept., 1965 in percent of percent of predicted employment Industry _12RI:2_, 1963

Manufacturing 103 98

Construction 137 127

Transportation, etc. 101 100

Wholesale Trade 100 104

Retail Trade 98 96

Finance and Real Estate 103

Service and Miscellaneous 105 104

All Industries 103 100 TABLE 5 MAJORPREDICTED OCCUPATIONAL EMPLOYMENTAND ACTUAL CLASSIFICATIONS INEMPLOYMENT SEPTEMBER, IN 1963 AND SEPTEMBER, 1965 BY (302 establishments in Newark Labor Actual employmentArea) icsmaLim Sept.,Employment 1963 Employment, Sept.,Predicted Actual 1965 centSept.,in Sept.,of 1965,employment 1963 in per- (percent) mentofActual predicted Sept., in percent 1965employ- (percent) Professional 8,090 8,244 8,566 106.102. 104. 99. ManagerialSemi-professional 6,3247,238 7,3596,536 614647,390 102. 100.100. SalesC!erical 271201 41888 271659 41971 271643 41837 101. 99. 97.98. SkilledSemi-skilled 14,91417,171 17139615,295 18,10815,053 105. 104. 98. UnskilledService TOTAL 102,02512,855 31344 104,611 131690 31411 104,724 131416 3,247 104.103. 97. 100. 94. in service occupations,where total employment inSeptember, 1965 was only94 percent of anticipated employment. For the other occupational groups, the differences werein no case greater than

4 percent.

Table 6 expands the datapresented in Tables 4 and 5 inorder

to show the variationswithin industry groups betweenactual and

predicted employment inSeptember, 1965. In a gross way these com-

parisons point to differences amongthe several industry groupsin

the ability of firms toforecast employment. At least they reflect

the degree of overall stability orvolatility in employment during

the period studied. As might be expected, constructionshowed the

greatest instabilityboth in overall percentages andfor the var-

ious occupational groups. Retail trade showed some declinein

actual employment, a trendwhich is the reverse of the actualchange

in this industry in theNewark Labor Area. However, as will be re-

called (see Table 4) our sampleof employment in this industry was

small--6.8 percent of employees fellinto the sample--and no doubt

even smaller inrespect to proportion offirms.

One may raise the questionof whether or not thelevel of ac-

curacy of employmentforecasting by firms was uniformfor the several

major occupational groups. In other words, was it tobe expected

that our respondents in givenindustry groups would tend to pre- than others? dict certain occupationalchanges with greater accuracy

A very rough measureof the variability in accuracyof prediction

across occupational groupsmay be derivedby calculating the average TABLE 6 PREDICTED AND ACTUAL EMPLOYMENT (302 establishments in the Newark Labor Market Area, September, 1963 to September, COMPARE6, BY MAJOR OCCUPATION AND INDUSTRY GROUPS 1965) Occupation facturin9 Manu- ** TIT struction Con- (2) portation(1) Trans- (2) 77)Wholesale Trade (2) (1) (7)Retail Trade and Real(1) FinanceEstate (2) MiscellaneousServices and(1) (2) ProfessionalSemi-professiona1 9997 9393 120152 106140 100 99 9898 107 83 107 83 9896 8896 121 92 116 91 110103 102110 ClericalManagerial 104 97 100 92 158126 120145 103100 102100 96 9893 101 99 101 97 107 92 105 91 100113 110 99 SkilledSales 107 99 101 96 131 60 129 60 101120 101120 105 93 107 91 113 103 91 92 103160 102150 110 94 104 92 UnskilledSemi-skilled 106107 105 97 140 77 126 68 99 9299 110 98 106120 122 110 96 96 7783 9377 120103 120100 Service TOTAL 103 91 8598 137--- 127___ 101 97 100 97 100 99 104100 102 100 98 96 103110 102110 105102 104 98 **(2) *(1) ActualEmploymen.t, Employment, September, September, 1965, 1965,in percent in percent of employment, September, of predicted employment for that date. 1963. - 13-

deviation of estimates. When this is done for the data in each column marked (2) in Table6, the following average de- viations are derived:

Ali Industries 2.4 Manufacturing 5.0 Construction 22.5 Transportation 4.2 Wholesale Trade 8.3 Retail Tiade 4.8 Finance and Real Estate 13.2 Service and Miscellaneous 6.2

Interpreted, these figures indicate thatthe predictions of sample firms in the transportation,communications, etc., industry group were distributed withconsiderable uniformity across oc.c.vpatioh the major smimmimp groups. On theother hand, firms in the construction industry tended towards extremedeviations in some of their predictions. Thus, for the construction industry as awhole

actual employment in September,1965 was 127 percent of predicted

employment. But the estimates rangel from145 percent for clerical

workers to 60 percent for sal(:s workers. To be sure, the index is

very gross; no attemptwas' made to weight the differences, although

a cursory examinationof the figures suggests that the rankingof

industries would not be different.

Predicted Expansion Needs

Analysis of findings to this point hasfocused on overall em-

ployment change in the Newark LaborArea as suggested by sample re-

turns from 302 establishments. The crucial question confronting manpower officials, however,is phrased in terms of the number of openings likely to occur in aparticular occupation or occupation group within a given period. Indeed, the question, framed in this way has been a guide to theorganization of training programs under the ManpowerDevelopment and Training Act.

What do the findings of the present surveydisclose with respect to the accuracy and usefulnessof the Occupational Training Needs

Survey as a guide to such decisions?

In Table 7 are shown the predictedjob openings expected to occur over the two yearperiod in major occupational groupsand

the actual change in employmentduring the same period. Results

are indicated forthe 302 establishments taken as awhole as well

as for the twoindustry groups for which sampling ratios were

reasonably high: manufacturing and transportation,communications

and public utilities,

As an overall guide to occupationalneeds in the Newark Labor

Area, the predictiJns might beconsidered reasonably useful, with

allowances made for the utilization ofother relevant manpower in-

formation. For example, sample establishmentsemployed more pro-

fessional workers in September,1965 than they had expected to em-

ploy. Demand for professionals was generallyknown to be vigorous.

Moreover, short-term public action atthe local level or even at

the level of the state is not capableof doing much to resolve dis-

closed shortages. To be sure a more accurateprediction in this

case by disclosing theactual level of competition might have a PREDICTED AND ACTUAL CHANGES IN NUMBER OF EMPLOYEES FOR SELECTED INDUSTRY GROUPS, BY MAJOR OCCUPATIONALTABLE CLASSIFICATIONS 7 BETWEEN SEPTEMBER, 1963 AND SEPTEMBER, 1965 ALL INDUSTRIES (302 establishments in the Newark Area) MANUFACTURING TRANSPORTATION, COMMUNICATIONS AND PUBLIC UTILITIES Occupation PredictedSept. '63 to Sept. '65Employment Change Actual PredictedSept. '63 to Sept. '65Employment Change Actual PredictedSept. '63 to Sept. Employment'65 Change Actual ManagerialSemi-professionalProfessional 212121154 476140152 166127 88 - -33 67 88 46 72- 7 1 SkilledSalesClerical 458381 - 83 442- 51139 264241288110 -181 890131 -3 202111 1 *15-24 2560 UnskilledSemi-skilledService TOTAL 2,586 225117835 2,699 937-561 97 2,060 739 37 1,256 485- 54 125 58 120 - 5- 7 - 16-

bearing on the actions of employers inrecruitment and in internal manpower planning and allocation.

The predicted increase in clerical workersin the sample as a whole was almostexactly matched by the actual increase in employment over the two-year period. This disclosure, emerging from the original survey, was thus capableof supporting the wide-spread impression of continuing shortagesin the clerical occupations. Somewhat difficult to explain but possiblydue to sampling inadequacy is the decrease in clericalemployees in menu-0

facturing. Although not included in the figures shownin Table 7,

it should be added that the major increase--justunder 600--in

clerical jobs occurred in the finance, real estateand insurance

industry group.

It is, finally, of interest to notethe sharp difference be-

tween predicted and actual employmentin semi-skilled jobs. The

difference may well be explained by theconjunction of forces

operatimg° in the labor market under conditionsof an upsurge in

demand. It appears that in the nation as awhole, the reduction

in unemployment rates over the periodunder study was particularly

marked for unskilled and semi-skilled workers. Skilled workers,

in short supply, are stretched asfar as possible, while expanded

production needs are met by increasing thenumber of operatives.

It will be noted from the figurespresented in Table 7 that the

major increase occurred inmanufacturing.

It is of particular interest toexamine the findings with

respect to selected occupationsknown to have been in relatively -17 -

short supply during the two years coveredby the present study.

The figures presented in Table8 indicate that, although in all cases some increase waspredicted, sample firms employed fewer draftsmen, automobile mechanics, and tooland die makers in

September, 1965 than they had two yearsearlier. On the other hand, as might be expected, the numberof programmers employed at the end of the periodsubstantially exceeded the estimate.

The Quality of Employer Response

The manpower survey with which the presentstudy is concerned

involved requesting employers toindicate employment in a given

list of occupations as of September,1963, and among other infor-

mation, to supply estimates of employment two yearslater, that

is, as of September, 1965. Understandably, in any sample of firms

some will fail to supplyall of the data requested. The question

which is crucial for the validity andusefulness of-the survey

being tested is the following: How many and what proportion of

firms surveyed failed to provide therequested two-year forecast?

The technical procedure followedin tabulating returns was to ex-

tend the September, 1963 figure toSeptember, 1965 whenever the

estimate was omitted and could notbe.,,secured by telephone or

follow-up. This procedure, if applied in a limitedproportion

of cases, is probably not serious; onecould reasonably assume

that the failure to forecastimplies that no change is expected.

However, if it is found necessary tomake such automatic extension TABLE 8 EMPLOYMENT CHANGES(302 establishments IN SELECTED OCCUPATIONSin NewarkSEPTEMBER, Labor Area) 1963 TO SEPTEMBER, 1965 Occupation September, 1963 Employment Predicted September, 1965 Employment Actual Draftsman 1,354 1,402 275 1,258 423 3 StenographerKeyProgrammer Punch Operatorc 48i258886 502895 566890 Auto1st Mechanic Class Metalworking 532 648544 660513 ToolMachinist and Die MakerMachine Hand 1,144 617629 1,305 655 1,280 560 - 19 -

of current figures in a high proportion of thesample, one could argue that an almost equally useful forecast could be derived by extrapolation of data for the universe derived from existing sources but adjusted for conditions in a givenlabor market area as these are understood by local EmploymentService officials.

As part of the test being attempted in the present study, the 302 establishments in the sample were classified inthree groups as follows:

1. In at least one occupational category an entry showing an expected increase or decrease was made in the column provided for estimating employment in September, 1965. This was taken to indicate that the respondent had made an attempt to forecast employment.

2. Entries appeared in the forecast column which were identical with employment figures for September, 1963, which could

be interpreted as implying no change; or, if no such entries appeared, there was some explanation indicating that the respondent

had considered the question of forecasting, but decided that no

change would occur; or, it was explained that no forecasting was

possible. -20-

explanation pro- 3. No entries appeared, nor was an vided; this is interpreted asimplying that respondents were unable or unwilling to supplythe requestedforecasts.*

In cases which fellinto categories two andthree, where no entriesappeared, editors repeatedthe employment figures 9 of sample es- for September,1963. The distribution in Table tablishments among the threecategories disclosessignificantly that this procedure was necessaryin 160 out of the 302 returns, In 60 cases no or over halfthe cases constituting thesample. clues were available toguide the decision thatemployment in

September, 1965 could withconfidence be assumed toparallel that of September,1963.

The classification offirms into the threecategories per- mitted some assessment of theassumption one mightreasonably

make that the larger firmswould more frequently havedeveloped

manpower plansand would thus morefrequently be found in cate-

gory one. The assumption is supportedin some industry groups

but not in all, atleast by the criterion ofmedian size of firm. size of firms in cate- In manufacturing,for example, the average

gory three issomewhat higher than in theother two. In the

edited the *In an undeterminednumber of cases, persons who returns communicated bytelephone with respondentsand determined that the absence offorecasts was meant to indicatethat no into cate- changes were anticipated. These cases would thus fall gory two aboverather than category three. NUMBER AND MEDIAN SIZECLASSIFIED OF EMPLOYMENT ACCORDING IN ESTABLISHMENTS TO WHETHER OR BYNOT INDUSTRY, PREDICTIONS WERE MADE TABLE 9 Expansion Needs (Newark LaborNo Expansion Area) Needs Industry or More OccupationsIndicated for One toIndicated Estimate but Ex Failurelained No Estimates Attempted Total 3 No. of Firms of EmploymentMedian Size No. of Firms of EmploymentMedian Size No. of Firms of EmploymentMedian Size 120 No. of Firms 143 Employment ' ConstructionManufacturing 66 8 115 24.5 39 4 113.5 50.5 38 1 ...... 13 43,077 883 Communications,Transportation, etc. lo11 197.5 85 21.517 81 ...... 25.5 2716 21,662 1,496 Finance,WholesaleRetail TradeInsurance Trade 18 3535 15 14 5 26 38 6,793 andServices Real Estate TOTAL 142 1613 131 85 100 26 3 2512 60 52 26 302 4718 102,025 1512,376 738 - 22 -

transportation and communications industry group the estab-

lishment with the largest number of employees (over 10,000)

was in category two, having indicated that noestimates by

occupation were available. Also, the one firm in category

three had almost 3,000 employees. In retail trade, the second

largest firm in the sample (over 1,000 employees) was in cate-

gory three. On the other hand, the largest firm in the sample

from this industry (3,500 employees) did supply estimates for

the target date two years later. In the service industries

and in finance, insurance and real estate, the largest firms

tended not to supply estimates that would serve the purpose

of manpower forecasting.

In summary, of the 14 establishments in the sample with

more than 1,000 employees, only six suppliedthe requested

estimates; eight were in category three, having failed to make

any response to this part of the questionnaire.

The conclusion appears warranted that the usefulness of

the survey as a means of predicting manpower needs was drastically

reduced by the failure of many firms, including many large ones,

, to supply considered responses to the request for forecasts of

job openings in the listed occupations. -23.-

Reasons for Inaccurate Forecast

In an effort to explore the reasonsfor differences between expected and actual employmentin September, 1965, employers were asked to check one or more ofthe following suggested causes whenever the figure theysupplied for the latter date exceeded or fell short of the expectednumber of employees:*

Reasons for differences betweenactual and expected:

A. Business poorer than expected B. Business better than expected C. Technological changes D. Qualified workers not available E. Insufficient information to make accurate prediction F. Other - please specify

If more than one reason, circle mostimportant.

The number of firms ascribing reasonsand distribution by type are shown in Table 10. Also shown are figures for the number of employees involved by broad industry group. It is worth noting that some 225 references were madeascribing differences in numbers employed to technological changes. Examination of the responses indicate thatoccupatiorsmost frequently involved were inthe

*The follow-up survey form, it will berecalled, reproduced the data supplied by respondents inthe original survey in 1963. Where no estimates had been made, theSeptember, 1963 employment figures were entered. Thus, our respondents who fell into categories two and three of theclassification discussed above were actually ascribing reasons for the notedchange in employment rather than the difference between "expected" andactual employment in 1965. REASONS FOR THE DIFFERENCE BETWEEN EXPECTED AND ACTUAL EMPLOYMENT TABLE 10 AS OF SEPTEMBER,facturing 1965 AND NUMBERManu- OF EMPLOYEES AFFECTED Construc-tion portation Trans- Wholesale Trade TradeRetail RealFinance Estate and Services Total NumberTotalTotal number ofnumber Em oflo of firmsees firms Affected reporting 143 90 13 8 1016 2712 3816 1813 4724 302173 A. B.Business Business poorer better than than expected expected -1335 -3281899 -23190 -4 -278 -52 64 87---6 -399 164 -3 -171 -523 762-31-34 -1220-1249 3189 4- It D.E.C. Qualified TechnologicalInsufficient workers informationchanges not available toaccurate make prediction (netincrease change) (+) -109 -18 -10 o +2-- -4-2 -32 -5 -4o -- -104 -16 -257 -81 +2 F. Other (net change) decreaseincreasedecrease (-) (+)(-) -278 -30 -33 13 -24 21 439 -365+473 108-83 -25-

unskilled and clerical classifications. Some reduction in employment was ascribed to the fact that qualifiedworkers were not available. Interestingly, the largest number, 104, occurred in the service industries. Examination of the returns suggests that the occupations inwhich this factor was revelant were: secretary, clerk-typist, and plumber-pipefit-

ter.

Furthermore, it may be noted that a small number of re-

spondents checked reason E--insufficient information to make

accurate prediction--although as has been seen, alarge num-

ber of firms failed in the original survey to offer any pre-

diction of employment two years !ater.

Finally, some references were made to other reasons for

differences. These included acquisition of a new company

through merger, and closing part of a plant. One respondent

noted that a construction project had just beenterminated.

Another referred to a "change in procedure," withoutindicating

its nature, as cause for a reduction in employment;finally,

one respondent ascribed areduction in service workers to con-

tracting out of maintenance functions. -26-

III

SUMMARY AND CONCLUSIONS

Accurate forecasts of manpower needs in locallabor mar- kets are indispensable to the Employment Servicein the perfor- mance of its primaryresponsibilities. But they serve equally pressing demands of educational agencies and thosecharged with planning economic development of areas; they also supplyinfor- mation needed by private firms concerned with manpowerand production planning and with in-plant training. The formulation and refinement of instruments capable of ensuring accurateshort- term and long-term projections thus becomeexceedingly important

tasks.

In this brief report an attempt is made to testthe efficacy and validity of measuring future manpower needsby means of an employer survey. .The survey follows from the reasonable assump-

tion that if one could amalgamate for a given marketemployer

estimates of future needs for workers in keyoccupations, one

could produce the necessary forecast ofdemand in the labor mar-

ket area as a whole. But the assumption is valid only to the ex-

tent that employers in a given market(1) are adequately sampled

and (2) are willing and able to perform the taskof estimating

future need. -27-

In the present study, employers whoresponded to a request for information in September,1963 were sent a follow- up form in theFall of 1965 on which was reproduced their estimate of employment as of September,1965. They were asked

to indicate actual employment in September,1965; and, where a discrepancy occurred between predicted and actualemployment,

to suggest an explanation for thediscrepancy.

The data discussed in the report suggest thatin overall

terms, estimates of changes in totalemployment in the sample

firms (302) were reasonably accurate. For all firms taken to-

gether the predicted and actual employment inSeptember, 1965

almost coincided. However, as might be expected, net changes

within industries and within occupations weremuch more variable.

Policies based on the estimates would havebeen quite sound in

respect to clerical occupations,would have overshot the mark in

skilled occupations, and would have fallensignificantly short

in the case of the semi-skilled occupational group.

In sum, one may conclude that theestimates collected in

the survey of September, 1963 might haveserved, in conjunction

with other information available forthe Newark Labor Market Area,

as a reasonably usefulinstrument of policy.

Nevertheless, study of the ultimate base ofthe estimates

leads to several reservations. In the first place, it seems

clear that in a highproportion of the returns, including anumber -28-

from companies with large work forces, the failure of company officers to estimate future needs made it necessary for the survey staff to project September,1963 data ahead for two years.

The findings of this study suggest that this did notprovide as solid a basis for occupational manpowerneeds as one might wish.

On the other hand, since no occupational breakdownof employment for the local area is available, it may be that this imperfect estimate is better than none.

A further difficulty arises in respect to the initial accuracy of the occupational dataprovided. Discussions with respondents in the course of the present study leave the impression that at least some of the figures supplied in the Fall of1963 were collected without precision. The request that actual figures be supplied for September, 1965 and compared with the earlier estimates for that date in some cases evoked the plea that thebasis for the occupational breakdown was not known, or was not clear to the person responsible for filling in the follow-upform. In several cases where respondents agreed to review the original occupational data, substantial revision appeared necessary.

The employer survey as a technique of projecting manpower needs by occupation is a costly instrument: however in the absence of alternative current sources of occupational data it may be necessary to use this technique. But the responsible agency must address itself to the need for enhancing its efficacy. The cooperation of employers is indispensable, expressed not only in at

29 -

completing a questionnaire, but in supplying estimates based on some considered judgmentof probable changes in employment and in occupational structure within the establishment. The data can be made more useful by ensuring that information supplied by employers is based on a uniform definition of occupations.

Employers concerned with the impact of changes in an area need to be convinced of the value of aggregate data for the area as a basis for their own planning. Perhaps the presence of employers on advisory groups would make it easier to promote such acceptance.

These improvements require more personal contact with respondents and would probably increase survey costs. To some extent these could be reduced by developing more rigorous yetsmaller samples. But even if the additional cost initially cannot be off-

set, the enhanced accuracy might make the effort worthwhile. APPENDIX A

Questionnaires NEW JERSEY STATEAffiliated EMPLOYMENT with United SERVICE States EmploymatDIVISION Service OF EMPLOYMENT SECURITYDepartment of Labor and Industry State of New Jersey TO BE RETURNED AREATO BUREAU SKILL OF SURVEY RESEARCH NEWARK LABOR BY OCTOBER 1, 1963 RS-152 (11-63) MARKET AREA AND STATISTIC SYMBOL CONFIDENTIALAftecial(Far Itassnscelaw) Snow Agency Use) WIII SECTION A. NUMBER OF EMPLOYEES ON ALL PAYROLLS ALL EMPLOYMENT ZUS PAYROLL PERIOD ENDING ON OR NEAREST SEPTEMBER 15, 1945PREDICTED FOR SEPTEMBERPREDICTED 15, 1948 FOR IZ:IZ TOTALSEPTEMBER 15, 1963 I FEMALE II TOT AL III FEMALE IV TOT AL V FEMALE VI I SECTION B. N AME (To be signed by person completing farm) TITLE HUMMER EXPECTED Z2coItus EMPLOY- SEPT. 1961 EMPLOYMENT CURRENT SEPT EMBER 1965 EXPANSION NEEDS SEPT EMBER 1968 SEPTEMBER REPLACEMENT1945 NEEDS SEPTEMBER 1968 SEPT- 191118-SEPT. 1984IN-PLANT TRAIN-TO COMPLETE ING IN Z....1-ILI OCCUPATIONAL TITLES MENT TOT AL FEMALE TOT AL FEMALE TOT AL FEMALE TOT AL FEMALE TOTAL FEMALE YEARSTIRO TEARS FIVE 1 II III IV V VI VII VIII IX X XI XII XXXX XIII xxxx XIV b.-- 3 Total LinesProfessional 4 - 12 65 LaboratoryElectronicsDraftsman orTechnician Chemist Assistant 98 AllProgrammerNurse, Other Practical Semiprofessional 121110 151413 Managerial XXXI XXXI RS-132 (11-63) (CONTINUED) TO BE RETURNEDAREA TO SKILL BUREAU SURVEY OF RESEARCH NEWARK AND LABOR STATISTICS MARKET AREA BY OCTOBER 1, 1963 SYMBOL (Follow Attached Instructions) (For State Agency Use)CONFIDENTIAL trW SECTION B. (CONTINUED) EXPANSION NEEDS REc-LA=EMENT NEEDS NUMBER TO COMP L ET E EXPECTED' IWZn10 OCCUPATIONAL TIT LES EMPLOY- SE196PT 1 EMPLOYMENT CURRENT SEPT EMB ER 1965 SEPT EMB ER 1968 SEPT EMBER 1965 SEPT EMB ER 1968 SEPT. 1965-SEPT. 1968 1N-PLANT TRA IN- ING IN Z7J. MENT TOT AL FEMALE TOT AL FEMALE TOTAL FEMALE TOT AL FEMALE TOTAL FEMA LE YEARSTWO YEARSFIVE 1716 Total Lines 17 - 25 11 III IV V VI VII VIII IX X XI XII XIII xtv 18 BookkeepingBookkeeper Machine Operator , 212019 SecretaryKeyClerk-Typist Punch Operator 242322 AllTabulatingStenographer Other Clerical Machine Operator 272625 Sales 283029 AutomobileTotal Lines MechanicBody29 - 57 Repairman 323133 CarpenterBrickBaker Layer 3534 CylinderCutter, Hand Pressman A RS-152 (8-63) (CONTINUED) TO BE RETURNED TO AREABUREAU SKILL OF RESEARCH SURVEY NEWARK AND STATISTICS LABOR MARKET AREA BY OCTOBER 1, 1963 SYMBOL (Follow Attached Instnwtions) (For State Agency Use) CONFIDENTIAL reW SECTION B. (CONTINUED) EXPANSION NEEDS REPLACEMENT NEEDS NUMBER EXPECTE D TO COMPLETE WZ3Xgo OCCUPATIONAL TITLES EMP LOY- SEPT.MENT1961 EMP LOYMENTCURRENT SEPT EMBER 1965 SEPTEMBER 1965 SEPTEMBER 1965 SEPTEMBER 1965 s SEPT. 1965-SEPT. 196 IN-PLANT TRAIN- ING IN :iZ 11 TOTAL 111 FEMALE IV TOTAL V FEMALE VI TOT AL VII FEMALE VIII TOTAL IX FEMALE X TOT AL XI FEMALE XII YEARS XIIITWO YEARSFIVE XII/ 40 Instrument Repairman 434241 MaintenanceMachinistLinotype Operator Mechanic, Building 464544 MeatMaintenanceOffset Cutter Pressman Mechanic, Factory 494847 RefrigerationPlumberOil Burner - Pipefitter Installer or Air-Conditioning Mechanic 525150 TailorStationarySheet Metal Engineer Worker 555453 AllWelderTool Other and Die Skilled Maker RS-152 (5-53) TO BE RETURNED TO BUREAUAREA OF SKILLRESEARCH SURVEY AND NEWARKSTATISTICS LABOR MARKET AREA BY OCTOBER 1, 1963 SYMBOL (Follow Attached Instrucrlons) (For State Agency Use) CONFIDENTIAL NUMBER EXPECTED 0 c5lare SECTION B. (CONTINUED) SEPT. CURRENT SEPT EMB ER EXPANSION NEEDS SEPTEMBER SEPTEMBER REPLACEMENT NEEDS SEPTEMBER IN-PLANTTO TRAIN- COMPLETE ING IN ZLaD2 OCCUPATIONAL TITLES EMPLOY- 1.4p,IT 1-61 TOTAL EMPLOYMENT FEMALE TOTAL 1965 FEMALE TOTAL 1965 FEMALE TOTAL 1965 FEMALE TOTAL 1965 FEMALE SEPT.. 1965-SEPT. 1961 YEARSTWO YEARSFIVE 57 :1 11 III IV V VI VII VIII IX X XI XII XIII XIV 605958 ChemicalTotalFireman, Lines Operator Stationary 59 - 66 Boiler 626163 RoutemanPresser, MachineHand 646665 AllSewing Other Machine Semiskilled Operator 696867 ServiceUnskilledCook 1 I EVALUATION OF OCCUPATIONAL TRAINING SURVEY No. RutgersInstitute of- The Management State University & Labor Relations in cooperation with To Be Returned To: RutgersInstitute of- The Management State University and Labor Relations NEWARK LABOR MARKET AREA Research Program (Follow Attached Instructions) CONFIDENTIAL NewDivisionDepartment Jersey of Employment Stateof Labor Employment and Security Industry Service New Brunswick, N.J. 08901 f Reasons for differences between actual and expected: CODE-COLUMN X 2xCOW SECTION A NUMBER OF EMPLOYEES ON ALL PAYROLLS ALL EMPLOYMENT A BBusiness Business poorer better than than expected expected Z3..]ZW PAYROLL PERIOD ENDING SEPTEMBER 15,1963ON OR NEAREST SEPTEMBER 15, 1965PREDICTED FOR SEPTEMBER 15 ACTUAL FOR FEMALE 1960 DC ETechnologicalQualified Insufficient workers information changes not available to make accurate prediction* , TOTAL I FEMALE II TOTAL III FEMALE IV TOTAL V VI CompanyF Other - pleaseIf more specify than one reason circle most important cr Name (To be signed by person completing form) Title 2DCOW SECTION B SEPT. TOTAL FEMALE TOTAL FEMALE REASONS FOR wz OCCUPATIONAL TITLES EMPLOY- MENT1961 SEPT. 15TOTAL 1963 SEPT.FEMALE 15 1963 V PREDICTEDSEPT. 15 1965 PREDICTEDSEPT. 15 1965 SEPT.ACTUAL 15 1965 SEPT.ACTUAL 15 1965 DIFFERENCES(PER CODE ABOVE) X I 1 Total Professional II III IV VI VII VIII IX 324 DraftsmanTotal Lines 4 - 12 5 Nurse,LaboratoryElectronics Practical or TechnicianChemist Assistant 10 All OtherProgrammer Semiprofessional 12 151413 Managerial (CONTINUED) EVALUATION OF OCCUPATIONAL TRAINING NEWARK LABOR MARKET AREA SURVEY __..____,_,...... m....mmuio , (Follow Attached Instructions) CONFIDENTIAL 3CC2COW SECTION B, (CONTINUED) SEPT. TOTAL FEMALE TOTAL FEMALE REASONS FOR -zwZ OCCUPATIONAL TITLES II EMPLOY- MENT 1961 III SEPT. 15TOTAL 1963 IV FEMALESEPT. 15 1963 V PREDICTED SEPT. 15 1965 VI PREDICTED SEPT. 15 1965 VII ACTUALSEPT. 15 VIII1965 ACTUALSEPT. 15 1965 IX DIFFERENCES (PER CODE ABOVE) X 1716 BookkeeperTotal Lines 17 - 25 201918 KeyClerk-TypistBookkeeping Punch Operator Machine Operator 232221 TabulatingStenographerSecretary Machine Operator 262524 SolesAll Other Clerical 282729 AutomobileTotal Lines Body29 - 57Repairman 323130 AutomobileBrickBaker Layer Mechanic 343335 CarpenterCylinderCutter, Hand Pressman (CONTINUED) EVALUATION OF OCCUPATIONAL TRAINING SURVEY NEWARK LABOR MARKET AREA (Follow Attached Instructiogs) CONFIDENTIAL Wre SECTION B. (CONTINUED) 03wzox OCCUPATIONAL TITLES EMPLOY- MENTSEPT.1961 SEPT. 15TOTAL 1963 SEPT.FEMALE 15 1963 PREDICTED SEPT. 15TOTAL 1965 PREDICTED SEPT.FEMALE 15 1965 SEPT.ACTUAL 15 TOTAL 1965 SEPT.ACTUALFEMZLE 15 1965 DIFFERENCESREASONS FOR(PER CODEABOVE) 36- Electrical Appliance Serviceman II III IV V VI VII VIII IX X 393837 InstrumentFirstElectrician Class Maker Metalworking Machine Hand 424140 MachinistLinotypeInitrument Operator Repairman 454443 MeatMaintenance Cutter Mechanic, FactoryBuilding 484746 PlumberOilOffset Burner Pressman - PipefitterInstaller 5049 SheetRefrigeration Metal Worker or Air-Conditioning Mechanic . 535251 ToolTailorStationary and Die Engineer Maker 565554 AllWelder Other Skilled EVALUATION OF OCCUPATIONAL TRAINING SURVEY NEWARK LABOR MARKET AREA (Follow Attached Instructions) CONFIDENTIAL w230)IdCr SECTION B. (CONTINUED) SEPT. 1961 TOTAL FEMALE .PREDICTED TOTAL PREDICTED FEMALE ACTUAL TOTAL ACTUALFEMALE REASONS FOR Iz7.i OCCUPATIONAL TITLES It EMPLOY- MENT III SEPT. 15 1963 IV SEPT. 15 1963 V SEPT. 15 1965 VI SEPT. 15 1965VII SEPT. 15 1965VIII SEPT. 15 1965 IX DIFFERENCES (PER CODEABOVE) X 5857 Total Lines 59 - 66 _ 616059 Presser,ChemicalFireman; Hand Operator Stationary Boiler .i 6362 RoutemanPresser, Machine c,9 666564 AllSewing Other Machine SemiskiHed Operator 696867 CookServiceUnskilled APPENDDC B

Employment By Occupation and Industry for September, 1963 APPENDIX B TABLE A EMPLOYMENT BY OCCUPATION AND INDUSTRY (302 establishments.in the Newark PREDICTED AND ACTUAL, SEPTEMBER, FOR SEPTEMBER, 1963 Labor Area) 1965 122.L.L63 Manufacturing Employment Pred. Act.Sept.,65 §v1.1L61 ConstructionEmployment Pred.sept.165 Act. Sept.163 Transportation Employment Pred. Act.Sept.165 EmploymentWholesale Pred. Act.Sept.165 Employment Retail Pred. Act.Sept.165 Professional 2518 26842645 24852451 113 15 123 17 171 18 683149 687153 676150 14 6 14 5 15 5 isaLla4380 4880 4277 ClericalSemi-professionalManagerial 537221.06 56362494 24945191 5562 6065 8778 27928156 81762798 81812864 287120 297117 276115 1354 773 1387 776 1346 779 Sales 2007 21177378 21387087 383 5 398 5 512 3 5460 76 5471 77 5520 91 196154 201151 182162 2455 328 2424 358 2224 370 UnskilledSemi-skilledSki:led 12377 81737090 12618 8912 13267 8658 228 22 256 25 322 17 3165 782 3186 840 3141 775 482229 238396 252473 289904 320905 866353 Service 591 627 536 1M AM 1M 1M AID 1M 392 392 381 8 8 8 530 540 540 Cook TOTAL 43077 25 45137 26 44333 26 883IM 1M 1M 949Imb MI 1208 1M 1M 1M 21662 7 21787 7 21782 3 1496 1M 1M 1M 1427 1488 6793 37 6877 39 6632 35 TABLE A (Contld) EMPLOYMENT BY OCCUPATION AND INDUSTRY FOR SEPTEMBER, 1963 (302 establishments in the Newark Labor PREDICTEDArea) ANDACTUAL, SEPTEMBER, 1965 12EILEI--Siiii.'65 gm19.17.naLFinance Pred. Act. !SP:It'll 12L:l65 EmploymentServices Pred. Act. ISP"'63 aPte'65 Employment Total Pred. Act. Professional 500554 523558 607506 46622559 46712572 26315196 63248090 65368244 64648566 ClericalManagerialSemi-professional 9123 788 9222 802 9718 728 2854 297 2881 307 2844 332 27201 7238 27659 7359 27643 7390 SkilledSales 127 28 128 31 131 46 1330 163 1361 166 1251 173 149i4 4888 15295 4971 15053 4837 Semi-skilledUnskilled 1174 30 1174 27 9939 25 1365 806 1365 824 1634 832 1285517171 1369017396 1341618108 CookService TOTAL 12376 45 7 12517 45 7 12727 50 7 15738 1667 35 15917 1734 36 16554 1614 47 102025 3233 111 104611 3346 115 104724 3129 118