On Unemployment Investigation in Small Areas

A C T A U N I V E R S I T A T I S L O D Z I E N S I S FOLIA OECONOMICA 152, 2000

III. APPLICATION OF STATISTICAL METHODS

C z e s l a w D o m a ń s k i * , K r y s t y n a P r u s k a *

O N U N E M P L O Y M E N T IN V E S T IG A T IO N IN S M A L L A R E A S

Abstract. The paper presents methods of small area statistics which can be applied to investigate the phenomenon of unemployment in given subpopulations depending on sampling and additional information about population. Some methods selected from this group are shown on the example o f analysis of unemployment among men, women and all inhabitants in the Łódź Macroprovince. The analysis is conducted on the basis of data collected in November 1996 in investigation of economic activity of population (Badanie Aktywności Ekonomicznej Ludności - BAEL-investigation).

1. INTRODUCTION

T h e ph enom enon o f unem ploym ent can be a serious prob lem fo r each co u n try ’s econom y both on the scale o f the entire coun try and on the scale o f a region. T he ratio o f jo b - seekers and econom ically active persons is one o f characteristics o f la b o u r m ark e t and allow s to assess the state o f econom y. D ifferent m ethods o f investigating unem ploym ent and its causes are developed. Small area statistics m ethods can be used for this purpose, too. They enable to analyse the problem o f unem ploym ent on a regional scale on the basis o f d a ta gath ered fo r the w hole p o p u latio n w hose p a rt is th e distinguished area or su b p o p u latio n determ ined in a way different th a n territo ria l assignm ent (e.g. a su b p o p u latio n o f wom en in a given p o pu latio n).

In this p ap e r we consider som e estim ators o f to tal value, w hich are used in sm all area statistics. T hey can be applied in unem plo ym en t inves­ tigations. T hen we determ ine the estim ates o f different quantities characterizing unem ploym ent (e.g. unem ploym ent rate) in th e whole Ł ó d ź M ac ro p ro v in ce

and in p artic u la r adm inistrative provinces separately for m en, w om en and fo r all in h ab itan ts.

2. SELECTED METHODS OF SMALL AREA STATISTICS

Small area statistics deals w ith m eth o d s o f using d a ta gathered in investigation representative for a given p o p u latio n (registers, c u rren t regis­ tra tio n an d o th er ad d itio n al in fo rm atio n ) in statistical analysis o f su b ­ p o p u latio n s o f a given p o p u latio n . G enerally, an assum ption is m a d e th a t relations occurring between param eters characteristic o f the entire p o p u latio n are also kept for the distinguished sub po p u latio n . In the co urse o f o u r consid eratio n s we assum e th a t the p o p u latio n is divided into H disjoint su b p o p u latio n s (sm all areas) an d G disjoint stratas.

It is assum ed th a t a p o p u latio n is investigated with respect to variable У on the basis o f random sam ple, and th a t som etim es in fo rm atio n on auxiliary variable X is available. M oreover, the follow ing d e n o ta tio n s for /i = l , ..., H and g = l, G are introduced:

N - nu m b er o f p o p u latio n elem ents,

N h - num b er o f p o p u latio n elem ents belonging to /i-th sm all area, N ' t - nu m b er o f p o p u latio n elem ents belonging to 3-th s tra tu m , N hg - n u m b er o f p o p u latio n elem ents belonging to h-th sm all area and

3-th stratu m ,

A - set o f indexes:{ 1, N ) ,

Aa - set o f indexes o f values o f У o r X observed in p o p u la tio n elem ents belonging to /i-th small area,

Лв - set o f indexes o f values o f У o r X observed in p o p u la tio n elem ents belonging to 3-th stratu m ,

A hg - set o f indexes o f values o f У o r X observed in p o p u la tio n elem ents belonging to h-th small area an d 3-th stratu m ,

n - num ber o f sam ple elem ents from the whole po p u latio n ,

nh - nu m b er o f sam ple elem ents from the whole p o p u latio n belonging to /i-th small area,

n g - nu m b er o f sam ple elem ents from the whole p o p u latio n belonging to 3-th stratu m ,

nhg - nu m b er o f sam ple elem ents from the whole p o p u latio n belonging to Л-th sm all area and 3-th stratu m ,

X - set o f indexes o f values o f У o r X d raw n for the sam ple from th e w hole p o p u latio n ,

Xh - set o f indexes o f values o f У o r X in p o p u la tio n elem ents belonging to h-th sm all area and which were found in the sam ple d raw n from the w hole p o p u latio n ,

A.„ - set o f indexes o f values o f У o r X in p o p u la tio n elem ents belonging to 0-th stratu m which were fou nd in the sam ple draw n from the w hole p o p u latio n ,

Xkt - set o f indexes o f values o f Y o r X in p o p u la tio n elem ents belonging to h-th small area and 3-th stratu m w hich w ere found in the sam ple d raw n from the whole p o p u latio n ,

y, - value o f У in i-th p o p u latio n elem ent, x, - value o f X in i-th p o p u latio n elem ent, T - to tal value fo r Y in the p o p u latio n i.e.

T = I У i (1)

i e A

Th - to tal value for У in h-lh sm all area i.e.

T» = £ Уi = E I У1 (2)

*6 Л.л, 0 = 1 ieA»,

Tg - to tal value for У in g-th stratu m i.e.

T „ = ľ Уi = £ Z У1 (3)

ígA .g h= 1 ieAhg

Thg - to tal value for У in /i-th sm all area and 0-th stratu m i.e.

T>* = (4)

(еЛл,

X , X h , X t , X hg - to tal values for X in the p o p u latio n , in the h-th small area, in the g-\h stratu m , also in th e h-th small area and in the g-th stratu m , respectively (form ulas for these values are analogous to form ulas

0 M 4 ) ) ,

Y_> X - m ean for У and X , respectively, in the p o p u latio n , Y h., X h - m ean for У and X , respectively, in Я-th sm all area, Y »’ ~ m ean f ° r ^ a n ^ X , respectively, in #-th stratu m ,

Yhe, X he ~ m ean fo r У and X , respectively, in h-th small area an d <?-th stratu m ,

^ %.> Ť g, Ťhg - estim ates for to tal values T, Th , T g, Thg respectively, ¥> Уь.> J.g, ¥hg - estim ates for m eans 7 , 7 h , 7 g, Y ^ respectively, x > x h., x'hg ~ estim ates fo r m eans X , Y h , X g, X ^ respectively. I he sim plest m ethod o f estim ation o f param eters o f variable d istrib u tio n for the small area is taking as a ran d o m sam ple for the sm all area elem ents oi the sam ple d raw n from the p o p u latio n which belong to this area, and estim atio n o f selected param eters on this basis.

E stim ato rs constructed in this way are called direct e s tim a to rs1. I hey could be characterized by large variance when sizes o f nÄr n A, n g tu rn ed o u t to be small in a given investigation. In o rd e r to ob tain m o re effective estim ato rs, som e additional in fo rm atio n on p o p u latio n and s tra ta is used to co n stru ct them . In this case indirect estim ators are o b tain ed .

We will now present selected estim ators o f to tal value o f Y defined for a sam ple o btained by one-stage stratu m sam pling (see. D o l 1991, p. 19-24).

T h e direct estim ators include the following: - H o rtv itz-T h o m p so n estim ator

Ťh. ( H T ) = I y j n t = I I y j n t (5) lei/,. s= 1 i ел/,,

- ra tio estim ator

E * <

% { R a l ) = • £ y , / n t (6)

ic;*. - co u n t ra tio estim ato r

Ťh{ C R a t ) = % = £ y j K t (7) £ 1бД*

i<úkni

(this is ra tio estim ato r in which x, = 1 o r auxiliary variable is c o n sta n t and tak es value 1),

- p o st-stratifica tio n estim ator

" I

( I * ,

. Y ^ y j n Ą (8)

I X j n t

- co u n t post-stratificatio n estim ator

o - i ß k - E * )

<9>

0=1 \ пЫ1 IeXhg J (this is estim ato r (8) for which x ; = 1).

In fo rm u las (5)—(7) n den o tes pro bability th a t i-th p o p u la tio n elem ent will be d raw n from a sam ple (often я , = n , / N , w hen i e A t ).

E stim atin g o f variance for the presented direct estim ato rs is n o t a sim ple task. It is estim ated on the basis o f a sam ple and it enables to co m pare the q u ality o f estim ators. M ean square e rro r defined by

M S E = E ( Ť , - T . ) 2 (10)

is also used for this p u rpo se, where d eno tes the w hole p o p u la tio n or this p a rt o f p o p u latio n which is exam ined in the investigation (e.g. sm all area). B oth estim ato r variance and M S E can be estim ated using different m ethods.

S ynthetic estim ators belong to the g ro u p o f indirect estim a to rs2. T hey are constructed w ith the assu m ption th a t relation s betw een an d X /у are the sam e in all p a rts o f sm all areas which belong to the sam e stra tu m , and the sam e as in a given stratu m i.e.

fo r 0 = 1, G and ft = 1, H.

W ith the above assu m p tio n the to tal value for У in h-th sm all area (ft — 1, Я ) can be presented by the follow ing form ula:

T„. - £ y, = i Tth = £ * 3 = £ ß„ X lh (1 2)

ieAk. g=l e = l -Aj e = l

T h e statistic o f the follow ing form :

t = i ß e ' X eh (13)

8=1

w here is the estim ato r o f p aram eter ßg, can be accepted as the estim ato r o f the to ta l value Th .

T h e synthetic ap p ro ach allow s to m odify direct estim ates (5)-(7). I f in form ula (13) we p u t ()g in place o f ß g, w here

K = Z yifai (14)

th en we will get the fo rm u la o f H orvitz-T h o m p so n estim ato r which is o f the form

T h e final form o f Ťh estim ato r results from the assu m p tio n th a t m ean value o f variable У in stratu m g i.e. У g is equal to Yhg i.e. m ean for the co m m o n p a rt o f /i-th sm all area and g-th stratum .

T h e form o f synthetic estim ato rs is sim ilar to form ulas o f post-stratifi- catio n. T h e difference betw een them lies in the fact th a t p o st-stratificatio n estim ato rs are fu n ctio n o f th ose elem ents from th e sam ple, w hich belong to the small area and synthetic estim ators are functions o f elem ents o f the ra n d o m sam ple from the whole po p u latio n . Let us notice th a t synthetic estim ato rs Ťh o f the to tal value Th for h = 1, ..., H have the follow ing p ro p e rty

(15)

T his results from the fact th a t in synthetic m eth o d we assum e

Z

Z

ie ka _ lel.e

X . g

(16)

W e m odify estim ato rs (6), (7) in a sim ilar way and we get - synthetic ra tio estim ator

C Z. УЧ'Н Ť„(Rat - Syn) =

Z

Z

Z

- synthetic co u n t ra tio estim ator

Ť„(CRat - Syn) =

Z

Z

0=1 • 0 0=1

(18)

F o rm u la s (5)—(9) and (13), (15), (17), (18) present estim ato rs o f the to tal value Th . In o rd e r to o b tain estim ators o f m ean У „ we need to divide Ťh by n u m b er N k .

In ad d itio n to the given estim ato rs fo r the sm all area in case o f stratu m sam pling o th er estim ators e.g. synthetic regression estim ates are k now n.

As well as synthetic estim ation we can also study regression estim ation w hose idea is different from basis o f co n stru c tio n o f syn th etic regression estim ato r (sec D o l , 1991).

T h e m eth od o f sam pling has an influence on the form o f estim ato rs used in investigations.

T h e presented estim ators are defined fo r one-stage s tra tu m sam pling.

3. METHODS OF ANALYSIS OF UMEMPLOYMENT IN SMALL AREAS

3.1. General remarks

1 he phen o m en o n o f u n em ploym ent can be analysed on the scale o f the w hole p o p u latio n (e.g. for a given co u n try o r a given g ro u p o f co u n tries) o r on the scale o f distinguished su b p o p u latio n s. In the latter case m eth o d s of small area statistics can be used to estim ate p aram eters characterizin g the lab o u r m a rk e t such as: n um ber o f econom ically active people, n u m b er of th e unem ployed, n u m b er o f people lookin g fo r a jo b d u e to: loss o f place o f w ork, leaving a jo b , in ten tio n o f tak in g u p a jo b after a b reak o r in ten tio n o f tak in g up the first jo b . T hese p aram eters m ay be treated as to tal values o f respective variables and estim ato rs presented in Section 2 can be used to estim ate them if sam pling was d o n e according to stratu m sam pling, individual and direct. In case we choose a differen t w ay o f sam pling, form ulas for estim ato rs o f to tal value tak e a different form and becom e m uch m o re com plex w hen the research is carried o u t on th e basis of in fo rm atio n gathered by m eans o f m ultistag e m eth o d s o f sam pling (see e.g. R u s s o , F a l o r s i 1992; B r a c h a , 1994; G o ł a t a , 1996).

A p p licatio n o f m eth o d s o f sm all area statistics to investigate th e p h e­ n om enon o f unem ploym ent in regional ap p ro ach is recom m ended because o f possibility o f using d a ta collected fo r research o f the w hole p o p u la tio n (both in representative studies and in com plete studies e.g. cu rrent registration o f jo b - seekers). C o n d u ctin g statistical research for large p o p u la tio n s is usually quite expensive and tim e - consum ing. M eth o d s o f sm all area statistics allow us to avoid carryin g o u t special research in o rd e r to collect the needed in fo rm atio n . It seems reasonable to follow the rule o f m axim um use o f the gathered statistical d a ta . D evelopm ent o f sm all area statistics provides m ore possibilities o f m ak in g analyses o n the regional scale.

Л he p ap e r presents analysis o f the problem o f u n em ploy m ent on the exam ple o f the Ł ódź M acroprovince, P oland . T h e m eth o d s applied here can be used in an analogous way for other regional research o f unem ploym ent.

3.2. Categories o f population by economic activity

In accordance with in tern a tio n al stan d ard s, p o p u latio n o f persons aged 15 o r m o re is divided into three categories:

- w orking, - unem ployed,

- econom ically inactive.

T h e w orking people and th e unem ployed are treated as econom ically active, while the rem aining p a rt o f p o p u latio n belongs to th e g ro u p o f econom ically inactive.

T h e analyses o f econom ic activity o f p o p u latio n are being carried o u t system atically in m any countries and they show various aspccts o f the problem . T his type o f research is also being developed in P o lan d (sec: K a ł a s k a , W i t k o w s k i , 1993) o n the basis o f e.g. cu rren t reg istratio n in em ploym ent agencies or special questionnaires. Investigation o f econom ic activity o f pop u latio n (B adanie A ktyw ności Ekonom icznej Ludności - BAEL) is one o f them , and it has been conducted since 1992. T h e in fo rm atio n is gathered fo u r tim es a year (in F e b ru a ry , M ay, A ugust and N ov em ber) on the basis o f a sam ple obtained in tw o-stage sam pling. T h e first stage units are d raw n , w ith stratificatio n according to adm inistrative provinces, which are, in tu rn , divided into rural stratu m and 2 to 5 u rb a n strata.

In the second stage units are households an d all the in h a b ita n ts o f the draw n household belong to the sample. A ccording with the definition accepted in B A E L -investigation (see: D o b r z y ń s k a , G a ł k a , K o s t r u b i e c et. al, 1996) all the people w ho w orked or n o t, in a given week b u t w ere em ployed by a certain em ployer, are included in to the g ro u p o f w orking.

T h e unem ployed are people aged 15 or m ore w ho d o n o t w ork b u t expect to sta rt w orking in the period o f the next 30 days o r fulfill th ree conditions:

- d u rin g the week un d er investigation they were no t w orking people, - actively looked fo r a jo b i.e. in the last 4 weeks proceeding the investigation they u n d e rto o k actions aim ed at finding a jo b ,

- were ready to take u p jo b in the week un der investigation o r the next one.

In o u r p ap e r we accepted a slightly different definition o f the unem ployed. T hey are persons w ho tre a t them selves as people looking for a jo b o r they are expecting to tak e up em ploym ent. T he o th er n o tio n s i.e. econom ically active, w o rk in g a n d econom ically inactiv e are u n d e rsto o d in th e way described earlier. H ow ever, it should be rem em bered th a t the change in the d efinition o f the unem ployed results in the change in n u m b er o f th e gro up s o f the econom ically active. T h e introduced m odifications were caused by th e in te n tio n to c a rry o u t an investigation d iffe ren t fro m th e B A E L analyses, w hose results are published by the C entral S tatistical Office in P o la n d , in spite o f the fact th a t o u r research uses statistical d a ta gathered

in B A E L -investigalion. T h e results o f o u r research offer a slightly different assessm ent o f un em ploym ent th a n the analyses based o n definitions accepted in B A E L -investigation.

3.3. Results o f research on unemployment in the Łódź Macroregion

T h e Ł ódź M acroregion (Ł ódź M acrop rovince) consists o f the Ł ód ź Province and the neighbouring provinces: P iotrków Province, P łock Province, S ieradz Province, Skierniewice Province.

D a ta gathered in B A E L -investigation in N ov em ber 1996 w ere used in the present study. T hey are show n in T ab . 1-5 in a way w hich is essential for o u r analysis. In o rd e r to determ ine estim ates o f to tal values o f som e variables which ch aracterize the la b o u r m a rk e t in the Ł ód ź M ac ro reg io n tw o estim ators were used viz.: direct estim ato r (9) an d indirect estim ato r (18). T his m eans th a t for sim plification o f calcu latio ns we used estim ato rs which correspond to one-stage stratu m sam pling (a p rovince w hich belongs to the Ł ó dź M acroregion was taken as a stratu m ). W e considered tw o sm all areas: su b p o p u latio n o f m en in the Ł ódź M acro reg io n an d su b ­ p o p u latio n o f w om en in the Ł ó d ź M acro reg io n .

T h e o btained results allow to fo rm u late conclusions co ncerning b o th the evalu atio n o f the used estim ato r and the situ atio n on the la b o u r m a rk e t in the Ł ó d ź M acroregion fo r m en an d fo r w om en ju d g ed from th e p o in t o f view o f econom ic activity o f p o p u latio n and causes o f loo king for a jo b . R esults o f the calculations are presented in T ab . 6-7.

Estim ates o f unem ploym ent rate obtained on the basis o f synthetic estim a­ to r (18) arc the sam e for the sub population o f m en and for the sub po pu lation o f wom en. T his results from a specific co n stru c tio n o f this estim ato r.

It is built with the assum ption th a t population structure (param eters which characterize the p opulation) does n o t differ significantly from th e stru ctu re o f small areas. We tak e th a t estim ato r (9) gives better estim ates because som e deviations from presented assu m ption can be observed. R esults from o th er studies o f the problem o f unem plo ym ent also su p p o rt this statem en t (see. D o m a ń s k i , N o w a k o w s k a , 1996; K o s t r u b i e с, K o w a l s k a , 1997).

T h e estim ates we o b tain ed p ro v e th a t th e re are m o re u nem p lo y ed w om en th an unem ployed m en b o th in the whole Ł ó d ź M ac ro reg io n and in p artic u la r adm inistrative provinces (the only exception is th e P io trk ó w Province). T h e m ain reason why people look fo r em ploy m en t is loss o f jo b , and the next one is th e in ten tio n to tak e u p th e first jo b . T h e low est percentage o f job-seekers are the people w ho are loo king for a jo b because they resigned from the previous one. T h e presented results show clearly th a t the situation o f w om en on the lab o u r m a rk e t in th e Ł ó dź M ac ro reg io n is less fav o u rab le th a n the situ atio n o f m en.

с

00

T a b l e 1 Structure of population of Poland and Łódź Macroprovince by economic activity in BAEL-sample (November 1996)

Specification

Number of persons

Totals economically active economically inactive

total men women total men women total men women

Poland 31 443 16 806 14 637 23 228 8 989 14 239 54 671 25 795 28 876 Łódź Province 882 462 420 733 274 459 1 615 736 879 Piotrków Province 558 301 257 392 154 238 950 455 495 Płock Province 470 253 217 324 119 205 794 372 422 Sieradz Province 374 197 177 241 98 143 615 295 320 Skierniewice Province 444 235 209 279 105 174 723 340 383 Łódź Macroprovince 2 728 1 448 1 280 1 969 750 1 219 4 697 2 198 2 499

S o u r c e : Author’s calculation on the basis of BAEL-sample from November 1996.

C ze sla w Do m ań sk i, K ry sty n a P rus ka

T a b l e 2 Structure of economically active population for Poland and Łódź Macroprovince in BAEL-sample (November 1996)

Specification

Number of cconomically active persons Number of unemployed unable to take up job in the period of

next 2 weeks

working unemployed

total men women total men women total men women

Poland 27 384 14 956 12 428 4 059 1 850 2 209 462 182 280 Łódź Province 752 398 354 130 64 66 12 6 6 Piotrków Province 486 263 223 72 38 34 7 2 5 Płock Province 412 230 182 58 23 35 3 2 1 Sieradz Province 342 185 157 32 12 20 5 0 5 Skierniewice Province 406 221 185 38 14 24 3 1 2 Łódź Macroprovince 2 398 1 297 1 101 330 151 179 30 11 19

S o u r c e : Author’s calculation on the basis of BAEL-sample from November 1996.

о VO On Un employment In v es tig at io n in S m all A re a s

Structure of number of the unemployed for Poland and Łódź Macroprovince in BAEL-sample (November 1996) 1

Specification

Number of people

Number of unemployed looking for a job awaiting to start a job

total men J women total men women total men women

Poland 3 981 1 798 2 183 78 52 26 4 059 1 850 2 209 Łódz Province 127 63 64 ₃ 1 2 130 64 66 Piotrków Province 71 38 33 1 0 1 72 38 34 Płock Province 57 23 34 1 0 1 58 23 35 Sieradz Province 32 12 20 0 0 0 32 12 20 Skierniewice Province 38 14 24 0 0 0 38 14 24 Łódź Macroprovince 325 150 175 5 1 4 330 151 179

S o u r c e : A uthor’s calculation on the basis of BAEL-sample from November 1996.

C ze sla w Do m a ńs ki, K ry sty na P rus ka

T a b l e 4 Number of unemployed awaiting to start job or ready to take up a job in the period of next two weeks for Poland and Łódź Macroprovince

according to reasons of looking for a job m BAEL-sample (November 1996)

Specification

Reasons for seeking employment

Totals loss of job resignation from job intention to come back

to work after a break

intention to take up the first job

total men women total men women total men women total men women total men women

Poland 1935 957 979 251 133 118 643 231 412 768 345 423 3 597 1 668 1 929 Łódź Province 66 33 33 11 5 6 25 10 15 16 10 6 118 58 60 Piotrków Province 39 20 19 2 1 1 9 6 3 15 9 6 65 36 29 Płock Province 38 17 21 2 0 2 6 2 4 9 2 7 55 21 34 Sieradz Province 15 7 8 2 2 0 3 0 3 7 3 4 27 12 15 Skierniewice Province 24 6 18 1 1 0 1 0 1 9 6 3 33 13 22 Łódź Macroprovince 182 83 99 18 9 9 44 18 26 56 30 26 300 140 160

S o u r c e : Author’s calculation on the basis of BAEL-sample from November 1996.

On Unem plo yme nt In v es tig at io n in S m al l A re a s

Structure of population of Poland and Łódź Macroprovince by sex in 1996

Specification Number of inhabitants (in thousands)

total men women

Poland 38 618.0 18 796.7 19 821.3 Łódź Province 1 113.3 514.9 598.4 Piotrków Province 643.7 314.8 328.9 Płock Province 521.8 255.0 266.8 Sieradz Province 412.8 203.2 209.6 Skierniewice Province 423.7 207.4 216.3 Łódź Macroprovince 3 115.3 1 495.3 1 620.0

S o u r c e : Rocznik Statystyczny 1997, GUS, Warszawa 1998, Tab. 8(152), p. 92.

C ze sla w D om ańs ki , K ry sty na Pru ska

Evaluations of total values for different variables characterizing labour market for Łódź Macroprovince determined on the basis of estimator (9)

Specification Number of economically active persons Number of unemployed

Number of persons who seek employment according to reasons of looking for a job

Unemployment rate loss of job resignation

from job

intention to oome back to work after a break

intention to take up the first job

men women men women men women men women men women men women men women

Łódź Province 323 215 285936 44 774 44933 23087 22466 3498 4085 6996 10212 6986 4085 0.139 0.157 Piotrków Province 208 262 170751 26292 22 590 13838 12624 692 664 4151 1993 6227 3986 0.126 0.132 Płock Province 173432 137187 15 767 22127 11654 13 276 0 1264 1371 2529 1371 4425 0.091 0.161 Sieradz Province 135 694 115935 8 266 13100 4822 5 240 1378 0 0 1965 2066 2 620 0.061 0.113 Skierniewice Province 143 350 118043 8 540 13 555 3660 10166 610 0 0 565 3 660 1 694 0.060 0.115 Ł ódź M acroprovince 983953 827 852 103 639 116 305 57061 63 772 6178 6013 12518 17 264 20320 16810 0.105 0.140

S o u r c e : A uthor’s calculation on the basis of BAEL-sample from November 1996.

On Un employment In v es tig al io n in S m al l A re a s

Evaluations of total values for different variables characterizing labour market for Łódź Macroprovince determined on the basis of estimator

SpedGcation Number of economically active persons Number of unemployed

Number of persons who seek employment according to reasons of looking for a job

"i

loss of job resignation from job

intention to come back to work after a break

intention to take up the first job

rate

men women men women men women men women men women шее women men women

Łódź Province 280 879 327084 41399 48 210 21018 24 476 3 503 4079 7962 9271 5095 5 934 0.147 0.147 Piotrków Province 184 891 193 210 23 857 24 930 12922 13504 663 6092 2982 3116 4 970 5194 0.129 0.129 Piock Province 151044 157 840 18 640 19478 12212 12762 643 671 1928 2015 2 892 3023 0.123 0.123 Sieradz Province 123 506 127 523 10 567 10911 4953 5115 660 682 991 1023 2311 2387 0.086 0.086 Skierniewice Province 127 230 132954 10889 11379 6877 7187 287 299 287 299 2579 2695 0.086 0.086 Łódź M acroprovince 867 550 938 611 105 352 114908 57982 63 044 5 756 6423 14150 15 724 17847 19 233 0.122 0.122

S o u r c e : A uthor’s calculation on the basis of BAEL-sample from November 1996.

C ze sla w Do m a ńs ki, K ry sty na Р гш к а

T h e o b tain ed results m ay be e rro r biased resu ltin g fro m th e use o f estim ato rs, which co rrespond to a sim pler sam pling th a n th e sam pling in B A E L -investigation.

4. FINAL REMARKS

M ethod s o f small a re a statistics give quite a wide rang e o f possibilities ol ap p licatio n in m an y fields o f econom ic and social research.

It seems they can have a m o re w idespread ap plication if there is a need to g ath er auxiliary in fo rm atio n necessary to build effective estim ato rs o f respective p aram eters in rep resentative or global research.

REFERENCES

B r a c h a Cz. (1992), Próba zbadania własności estymatorów wartości globalnej dla dziedzin w losowaniu dwustopniowym, „Wiadomości Statystyczne”, 11, 9-13.

B r a c h a Cz. (1994), Metodologiczne aspekty badania małych obszarów, „Z Prac Zakładu B adań Statystyczno-Ekonomicznych. Studia i M ateriały” , 43 (Warszawa).

D o b r z y ń s k a J., G a ł k a C., K o s t r u b i e c S. et al. (1996), Aktywność zawodowa i bezrobocie tv Polsce w listopadzie 1995 r., [w:] Studia i analizy statystyczne. GUS, Warszawa. D o 1 W. (1991), Small Area Estimation. A Synthesis Between Sampling Theory and Econometrics,

Wolters-Noordhoff, Groningen (The Netherlands).

D o m a ń s k i Cz., N o w a k o w s k a В. (1996), Women 's Labour Market in Poland in Transition, „Polish Population Review” , 9, 131-146.

(j o ł a t a E. (1996), Statystyka małych obszarów w analizie rynku pracy, „Wiadomości

Statystyczne” , 3, 45-60.

K a ł a s k a M., W i t k o w s k i J. (1993), Program badań w zakresie statystyki pracy, „Wiadomości Statystyczne” , 11, 6-8.

K o s t r u b i e c S., K o w a l s k a A. (1997), Efektywność polityki rynku pracy. Studia i analizy statystyczne, GUS, Warszawa.

R u s s o A., F a l o r s i P. D. (1992), Conditional and Unconditional Properties o f Small Area Estimator in Two Stage Sampling, Paper presented on the conference “ Small Area Statistics and Survey Designs”, Warsaw, September.

S c h a i b l e W. L. (1993), Use o f Small Area Estimators in U.S. Federal Programs in Small Area Statistics and Survey Designs. International Scientific Conference, Warsaw, 30 September - 3 October 1992, Vol. 1, Invited papers, Central Statistical Office, Warsaw, 95-114.