55

Introduction

Identifying and measuring poverty is an im-portant issue facing policymakers in the wholeEuropean Union. In March 2000 the Lisbon Eu-ropean Council agreed on the need to take stepsto make a decisive impact on the eradication ofpoverty by 2010. In December 2000, at the NiceEuropean Council, the Heads of the States andGovernments implemented a decision that theefforts against poverty and social exclusion mightbe best achieved by means of the open method ofco-ordination. Key elements of this approach arethe definition of commonly agreed objectives forthe European Union (EU) as a whole, the devel-opment of appropriate national action plans to

meet these objectives, and the periodic reportingand monitoring of the progress made.

Fight against poverty was launched long be-fore Lithuania’s integration into the EU. In 2000,Lithuania has developed a Strategy on PovertyReduction, which was used for the elaborationand approval of Program on Implementation ofPoverty Reduction Strategy. Lithuania joined theCommunity process of reduction of poverty andsocial exclusion in 2002 with the signing of memo-randum of agreement with the European Com-mission. Following the document and the provi-sions of Accession Partnership, the governmentof the Republic of Lithuania backed by the Eu-ropean Commission worked out the Joint Inclu-sion Memorandum, which established the main

DETERMINANTS OF POVERTY IN LITHUANIA

Algimantas MISIÛNAS

Mykolas Romeris UniversityAteities g. 20, LT-08303 Vilnius;

Vilnius UniversityUniversiteto st. 3, LT-01122 VilniusE-mail: algis.misiunas@gmail.com

Graþina BINKAUSKIENË

Department of Statistics to the Government of the Republic of LithuaniaGedimino av. 29, LT-01500 Vilnius, Lithuania;

Vilnius Gediminas Technical UniversitySaulëtekio al. 11, LT-10223 Vilnius, Lithuania

E-mail: gb@ygle.lt

Abstract. This article examines the determinants of poverty in Lithuania. This study is based on 2006Household Budget Survey data.

Regression analysis technique was used to identify the contributions of different variables to pov-erty. It was attempted to explain the level of expenditure per capita – the dependent variable – as afunction of a variety of explanatory variables. The independent variables such as size of the household,living place of the households and various household heads characteristics: age, sex, education, socio-economic status were included into regression model.

JEL I320, I380.Keywords: poverty, Lithuania, determinants, household.Reikðminiai þodþiai: skurdas Lietuvoje, skurdà lemiantys veiksniai, namø ûkis.

ISSN 1822-8011 (print)ISSN 1822-8038 (online)

INTELEKTINË EKONOMIKAINTELLECTUAL ECONOMICS

2007, No. 1, p. 55–63

VE

RI

TIATAS

IVS

TI

56

challenges in fight against poverty and social ex-clusion and outlined the main measures of thatpolicy. By signing this document, Lithuaniapledged to initiate the inclusion of joint EU ob-jectives for fight against poverty and social exclu-sion into the national policy and defined the mainspheres of policy, which should be monitored andcontrolled in future.

The World Bank (1990) defined poverty as“the inability to attain a minimum standard ofliving”. Lipton and Ravallion (1995) stated that“poverty exists when one or more persons fallshort of a level of economic welfare deemed toconstitute a reasonable minimum, either in someabsolute sense or by the standards of a specificsociety”. Poverty in the modern society usually isdefined as a lack of income and other resources(financial, cultural and social) guaranteeing a tol-erable standard of living to the people of the state.From these definitions we can say that the con-ception of poverty is multimeaningful, relative andis subject to the standard of living in the country.

Referring to that the Laeken EuropeanCouncil in December 2001 endorsed the first setof 18 common statistical indicators for social in-clusion which will allow comparable monitoringof the Member States progress towards the agreedEU objectives. These indicators cover four impor-tant dimensions of social inclusion: financial pov-erty, employment, health and education.

Statistics Lithuania in their analysis of pov-erty is analyzing the evolution of poverty levelsand poverty profiles using these defined Laekenindicators. This kind of analysis focuses on whathappening to poverty during the time periods aswell as what happening to the composition of thepoor according to several demographic and so-cioeconomic characteristics. This knowledge canbe useful since it allows us to know whether pov-erty is increasing or decreasing as well as thechanges in the composition of the poor.

However, it does not provide us with muchinsight about the causes of poverty. For instance,we know that poor people tend to have low levelsof education; but are they poor because they havelittle education, or do they have little educationbecause they are poor? On the other hand whyrural people have low levels of education in thefirst place: Were the school fees too high? Wasthere no school nearby? Was the quality of theeducation abysmal? Were their parentsunsupportive, or even hostile to education?

For Lithuania, while there has been a lot ofworks on a descriptive analysis of the characteris-tics of the poor, to my knowledge, there is nopapers to an empirical modeling of the determi-nants of poverty using nationally representative

data. In this article, I have tried to extend thedescriptive analysis of the Lithuania poverty bymodeling the determinants of poverty, using datafrom the 2006 Household Budget Survey.

This paper is organized as follows. The nextsection describes measures of poverty approachto modeling the determinants of poverty. Section3 introduces the HBS data set and survey meth-odology and introduces the set of main conceptsused in this analysis. Section 4 presents the re-gression model, explains dependant and explana-tory variables. Section 5 presents empirical results.

1. Measuring poverty

To measure the pervasion of poverty or pov-erty determinants it is necessary to be able to dis-tinguish the poor from the non-poor. The tradi-tional approach involves establishing an incomethreshold and calculating how many individuals,families or households fall below it. The questionis how to establish the income threshold. There isno single correct approach; a wide range of meth-ods has been used in different countries and atdifferent times. Moreover, there is the questionof whether income itself is a reliable indicator ofliving standards.

Most modern definitions of poverty using anincome threshold set at a particular fraction ofmean or median income. This approach has beenused by international bodies such as the Euro-pean Union, World Bank and the OECD.

For this study a little bit different approachwas used: instead of income consumption expen-diture was used as the poverty indicator. The de-cision to use consumption expenditure is ex-plained in the section 4.

2. Data source

Data source used for research was providedby Statistics Lithuania and contains HouseholdBudget Survey data for the year 2006.

Survey Methodology. The target population ofthe HBS is based on private households inLithuania. Persons living in the institutionalhouseholds (nursing homes for elderly people,imprisonment institutions, compulsory militaryservice installations, etc.) have been excluded fromthe current survey. The household and its mem-bers can be classified according to various socio-economic and demographic characteristics such asincome, education, place of residence, type of em-ployment.

Reference Periods. The selected households arebeing surveyed for the period of one month. Af-

Algimantas Misiûnas, Graþina Binkauskienë

57

ter one month other households replace them.There were different reference periods for thevariables included in the Surveys. For the socio-demographic variables the reference period wasat the moment of the interview. For the incomeand non food expenditures variables, the refer-ence period is one month, for food expendituresreference period is two weeks.

Geographic Coverage. The Survey is statisticallyrepresentative at the national level, at the urbanand rural level and at the counties level as well.

Sampling Design. The population register wasused as a sampling frame. Stratified sample de-sign with simple random sample and two-stagecluster sample was used in strata. All Lithuanianterritory was divided in 31 not overlapping groups– strata. The biggest towns of Lithuania counties,medium, small towns and rural areas of countiesare divided in separate strata. Sample of house-holds was selected from each stratum. Differentsample design was used in each stratum: a simplerandom sample of persons 16 and older is drawnfrom the Population Register in major countytowns and two-stage cluster sample design wasused in the medium and small towns of countiesas well as in rural areas of counties.

Main concepts used in the analysis

Household – is an association of people tiedwith relationship or other personal bonds whohave common budget, have meals together andare accommodated in one housing unit (an apart-ment, etc.). One person may also comprise ahousehold.

Household head – is a person who has the high-est income. Due to the fact, that household mem-bers’ income may differ by different months, thehead of the household is a person, who, based onthe opinion of the household members, receivesthe highest yearly income. In cases, when it is im-possible to determine a person receiving the high-est yearly income (for e.g. all the family membersare involved in farming activities and the incomereceived by the family may not be assigned to oneof the members), the head of the household shallbe determined by the family itself.

Social-economic group is determined based onthe income source of the head of the household.Based on this attribute, the following householdgroups are pointed out:

• self employers in agriculture – households,where the income by the head of the householdis received from individual agricultural activities;

• hired workers – households, where the in-come of the head of the household is receivedfrom hired work within public or private sectors;

• self employers – households, where the in-come of the head of the household is receivedfrom business, trades and free professional activi-ties;

• pensioners – households, where the incomeof the head of the household is received in theform of pension;

• other – households, where the income ofthe head of the household is received in the formof various benefits, scholarships, income fromproperty, as well as other income sources.

Household type is determined based on de-mographic structure of the household. Householdtypes are as follows:

• Single person;• Single adult with children under 18;• Couple without children under 18;• Couple with children under 18;• Other households with children under 18.

This type of a household includes householdsconsisting of parents with children under 18 andolder, households of few generations with chil-dren under 18, grandparents with grandchildrenunder 18, etc.

• Other households without children under18.

Consumption expenditure includes expenditurein cash and kind for household consumptionneeds, i.e. food, clothes, dwelling, health care,culture and recreation.

3. Empirical model

Tabulated or graphical information on thecharacteristics of the poor is remarkably helpfulin painting a profile of poverty. However, it isnot always enough when one wants to discoverthe relative contributions of different influenceson poverty.

The most widespread technique used to iden-tify the contributions of different variables topoverty is regression analysis. This kind of analy-sis attempts to explain the level of expenditure(or income) per capita – as a function of a varietyof explanatory variables.

A regression estimate shows how closely eachindependent variable is related to the dependentvariable (e.g. consumption per capita), holdingall other influences constant.

Having data available from Lithuanian HBS,it is possible to formulate an empirical model fortesting the determinants of poverty in Lithuania.

Determinants of poverty in Lithuania

58

Dependent variable

Lithuania HBS collects information abouthousehold disposable income as well as the in-formation about household consumption expen-diture. Before starting the analysis, it is necessaryto decide which indicator will be used as a wel-fare indicator in poverty analysis. Most of thecountries prefer to use consumption expenditureinformation for the following reasons:

– In countries, where agricultural income isquite important source of income, income oftenis very lumpy. Farming households receive a largeamount of cash income in summer and autumn,and receive very little the rest of the year. On anincome basis, a household which most would viewas wealthy may be categorized as poor if the in-terview of that household was done at the begin-ning of the year or the oppositely, poor house-hold can be treated as wealthy, if the interviewwas conducted at the moment when householdreceived income for the sold harvest. At this pointof view expenditure is a smoother measure ofwelfare through time.

– Data on expenditures are generally morereliable and stable than income data. Householdsare often more willing to truthfully report theirconsumption expenditure than their income.

All these comments are suitable and forLithuanian HBS data.

In the Lithuanian Poverty Reduction Strat-egy when analyzing the prevalence of poverty andplanning the measures of social and economicalpolicy the main criterion of poverty was selectedthe relative poverty line estimated as 50 per centof the average consumption expenditure.

Concluding all listed above comments and tobe in one line with Lithuanian Poverty Reduc-tion Strategy it was decided to include per capitaconsumption expenditure as the dependent vari-able.

Independent variables

Many variables can be considered as the de-terminants of wealth, and thus, of poverty. Theset of independent variables that are hypothesizedto be determinants of consumption includeshousehold and household head characteristics.The key selection criteria for these variables wasexogeneity. As the goal of the model was to infercausality, variables which might be affected bycurrent consumption in a household – endog-enous variables – are excluded from the model.Selection of potential determinants was guidedby the results of the poverty profile conducted inthe previous studies.

Endogenous, or jointly determined variables,have values which are determined through thejoint interaction of other variables within thespecified system (Judge et al. 1988, p. 601). Incontrast, exogenous variables are variables thataffect the levels of the endogenous variables, butwhose own levels will be determined outside thesystem. Exogenous variables are assumed to in-fluence the values for the endogenous variables,but are not influenced by those variables in re-turn because no feedback relation between theendogenous and exogenous variables is assumed.Examples of endogenous variables that are likelyto be an outcome of current household living stan-dards (as measured via consumption levels) arethe possession of durable goods by householdmembers, dwelling characteristics, current schoolattendance of children in the household, and soon.

Endogenous variables are not selected as re-gressors because they are determined by currenthousehold living standards, i.e. by income or con-sumption expenditure.

So, the objective, selecting independent vari-ables, was to select variables whose values deter-mine the level of household welfare, but they cannot be explained be household welfare measure.

The set of regressors, or independent vari-ables, that were chosen as possible determinantsof poverty in Lithuania are as follows:

Characteristics of the household:• size of the household,• number of children under 7 years in the

household,• number of children 8–17 years in the house-

hold,• number of members with highest education

in the household,• number of dependants in the household• number of earners in the household• household type (reference group – single

person household)• living county of the household (reference

group – Vilnius county)• living in rural or urban area (reference group

– urban)

Characteristics of the household head:• age of the household head,• sex of the household head,• marital status of the household head (ref-

erence group – single),• education of the household head (refer-

ence group – highest education),• socio-economic status of the household

head (reference group – others).

Algimantas Misiûnas, Graþina Binkauskienë

59

Finally regression equation, as applied to thispoverty analysis, look likes this:

where z is the poverty line, yi is per capita con-sumption expenditure, the are the “explana-tory” variables and the are the coefficientsthat are to be estimated and is a normally dis-tributed random error term.

Following most similar studies, semi-logarith-mic form of the regression was choose. This in-troduces some non-linearity into the model, andtypically improves goodness of fit measures incomparison with similar estimations based on theabsolute value of consumption. This model wasestimated for the whole country and separatelyfor urban and rural areas.

4. Empirical results

Table 1 presents the parameter estimates ofthe regression model for the determinants of pov-erty.

The size of the coefficient for independentvariable gives you the size of the effect that vari-able is having on dependent variable, and the signof the coefficient (positive or negative) gives thedirection of the effect.

Since dependant variable is in natural loga-rithm form the estimated coefficient tells approxi-mate expected percentage increase (if the coeffi-cient is positive) or decrease (if the coefficient isnegative) of the ratio of consumption expendi-ture per capita and poverty line when that inde-pendent variable increases by one, holding all theother independent variables constant. In case ifindependent variable is classification variable –set of dummy variables are created and includedin the model. Last dummy variable (called refer-ence) always excluded from the model to avoidlinear dependency of independent variables. Inthis case estimated coefficient tells approximateexpected percentage increase (if the coefficientis positive) or decrease (if the coefficient is nega-tive) of the ratio consumption expenditure percapita and poverty line when comparing to thereference group dummy, holding all the otherindependent variables constant.

The fit of the fixed effects model for the wholecountry is estimated with an R2 of 0.341 with asample of 7,178. The R2 for the rural model was0.320 with a sample of 2,844 households. For theurban model the R2 was 0.349 for a sample of 4,334households.

Impact living place. Household’s living in ru-

ral areas have negative effect on welfare compareto urban household by -0.135. As showed the re-sults of regression, to living in any county inLithuania (except Klaipëda) have negative effecton household standards of living. The highestnegative effect was found in Utena and Alytuscounties. The only Klaipëda county showed apositive effect on households welfare compare toVilnius county analyzing regression model for thewhole country. Analyzing regression model forurban areas the same picture was found, but ana-lyzing regression for rural areas, the all countiesshowed negative effect compare to VilniusCounty.

Impact of size and type of the household. Asit was already expected, household size has a nega-tive effect on households’ well-being: the biggerhousehold, the worse well-being of the household.Effect of household size is smaller in rural areasthan in urban. Any composition of the householdcompare to the single person household also havenegative effect on households’ welfare. Multigen-erational households (with or without children)have the highest negative effect compared tosingle person households. All these conclusionscan be said about urban households as well. Inrural areas situation is a little bit different. Liv-ing as a couple (with or without children) has posi-tive effect on household well being in rural areas,even single parent households are better-off com-pare with single person households in rural areas.Just multigenerational households have a nega-tive effect on welfare in rural areas.

Impact of number of dependants and num-ber of earners in the household. Examining thenumber of dependants, it can be noticed that thisindicator have negative impact on household well-being. The more dependants in the household,the poorer household living conditions. The op-posite situation is analyzing number of earners inthe household. Contrarily, number of earners havea positive impact on household welfare: moreearners in the household, the higher living of stan-dard in the household.

Impact of number of children in the house-hold. The number of children in the householdhas a negative impact on the welfare of the house-hold. Households with more children, holdingother variables constant, will tend to be poorerthan households which have fewer children. Num-ber of children under 8 years have higher nega-tive impact that number of children between 8and 18 years old. Surprisingly, number of childrenhave positive effect on households welfare in ru-ral areas. We were able to found the only one

Determinants of poverty in Lithuania

60

a Dependent Variable: Logarithm of Consumption expenditure and poverty line ratiob Weighted Least Squares Regression – Weighted by sampling weight* significant at 1 percent level.

(Constant) 1.479* 593.2 1.665* 585.3 0.761* 149.4Living in rural area -0.135* -190.7Living in Kaunas county -0.088* -119.6 -0.017* -20.6 -0.326* -207.9Living in Klaipëda county 0.036* 38.7 0.059* 58.8 -0.132* -61.1Living in Šiauliai county -0.060* -58.2 -0.024* -20.5 -0.267* -130.6Living in Panevëþys county -0.164* -137.6 -0.138* -92.3 -0.254* -128.9Living in Alytus county -0.239* -160.6 -0.248* -142.8 -0.275* -100.3Living in Marijampolë county -0.130* -80.4 -0.123* -57.6 -0.221* -91.9Living in Telðiai county -0.043* -26.4 -0.064* -33.6 -0.068* -23.3Living in Tauragë county -0.206* -99.3 -0.261* -84.0 -0.301* -110.9Living in Utena county -0.239* -147.9 -0.209* -100.9 -0.366* -146.7Size of Household -0.199* -138.1 -0.193* -114.9 -0.355* -132.9Type of Household

Single adult with childrenunder 18 -0.121* -68.4 -0.184* -87.4 0.049* 15.5Couple without children -0.103* -70.1 -0.203* -116.3 0.111* 41.2Couple with children under 18 -0.204* -125.8 -0.290* -149.5 0.012* 4.1Other households withchildren under 18 -0.189* -104.1 -0.264* -124.5 -0.014* -4.0Other households withoutchildren under 18 -0.176* -118.8 -0.204* -116.9 -0.062* -23.2

Number of dependant members in thehousehold -0.002* -3.6 -0.011* -16.0 0.042* 34.2Number of earners in the household 0.086* 116.8 0.043* 51.3 0.246* 162.7Number of members with highesteducation in the household 0.114* 171.4 0.107* 149.5 0.241* 129.3Number of children under 8 years in the HH -0.043* -46.7 -0.041* -35.8 0.018* 11.5Number of children 8–18 years in the HH -0.005* -6.3 -0.001 -1.6 0.067* 47.5Characteristics of Head of HouseholdGender of HH Head-Female 0.000 0.1 0.004* 5.5 -0.076* -60.5Age of HH Head (years) -0.006* -195.4 -0.006* -168.4 -0.002* -34.7Education

College -0.106* -90.2 -0.119* -92.4 0.012* 4.0Secondary education -0.203* -176.4 -0.197* -157.5 -0.088* -30.9Basic education -0.390* -260.2 -0.433* -242.5 -0.126* -40.0Primary education -0.389* -234.1 -0.387* -181.9 -0.298* -89.8

Socio-economic groupSelf-employed in agriculture 0.059* 23.9 0.294* 24.0 0.254* 70.1Employees 0.037* 20.6 -0.047* -22.0 0.222* 67.7Self-employed, employers 0.373* 167.4 0.121* 45.8 1.008* 246.9Pensioners -0.091* -46.6 -0.188* -78.7 0.045* 13.6

Marital Status:Married 0.275* 187.4 0.325* 189.5 0.174* 62.8Cohabited 0.147* 90.6 0.255* 134.3 -0.018* -5.8Spouse lives separately 0.041* 18.0 0.130* 42.9 0.131* 38.4Widowed 0.093* 63.9 -0.003 -1.6 0.310* 121.1Divorsed -0.034* -24.7 -0.008* -5.1 -0.057* -21.5

Table 1. Model of the determinants of poverty

RuralUrbanOverall

Coefficients T statistics Coefficients T statistics Coefficients T statistics

Algimantas Misiûnas, Graþina Binkauskienë

61

reason for that: households with children receivesbenefits form the government and sometimes it isthe only source of income. Therefore these house-holds with children compare to other householdsare better-off. Coefficient for number of childrenbetween 8 and 18 years old in urban areas is notstatistically significant.

Impact of Age of Household head. As It couldbe seen from the table above, the age of thehousehold head has a relatively small impact onthe welfare of the household. However, it is im-portant to note the high level of statistical sig-nificance of the coefficients for the whole coun-try and for rural and urban areas separately.Households headed by older individuals, holdingother variables constant, will tend to be poorerthan those headed by younger individuals.

Impact of Gender of Household head. Coef-ficient for the gender of the head of household isnot statistically significant for the whole country,but is statistically significant in urban and ruralmodels. Female-headed household have negativeeffect compare to male households in rural areas.In urban areas the effect of the female headedhousehold on the household welfare is positive,but quite small just 0.004 compare to male-headedhouseholds.

Impact of Education of Household head. Thecoefficients for the variables for education ofhousehold head are very significant. The follow-ing conclusion can be done from these coefficients:the higher education of household head, the lowerdecrease of welfare. If head of household havethe primary education, there is a substantial de-crease of household well-being: approximately –0.389 for the whole Lithuania and of – 0.387 and– 0.298 respectively in urban and rural areas com-pared to the households which head have the high-est education level. While for the householdswhich head have a colleague degree decrease ofwelfare is much lower –0.106 for the wholeLithuania and in urban – 0.119 compared to thehouseholds which head have the highest educa-tion level. In rural areas colleague degree have apositive impact on household welfare compareto highest education.

Impact of Socio-Economic group of House-hold head. The coefficients on socio-economicstatus of household head mostly are positive; allare statistically significant for the whole countryand in both rural and urban areas. Self employ-ment in agriculture as well as self-employment oremployer in non agriculture of household head

have positive effect on the welfare compare towhose households which head have so call “other”socio-economic status. However, what is surpris-ing is the negative coefficient for household whichhead is employees in urban areas. These house-hold have negative effect at –0.047. Negative ef-fect on welfare was observed in the pensioners’household as well. But this finding is commonjust for the urban areas.

Impact of marital status of Household head.The coefficient for the marital status of house-hold head surprisingly are statistical significant,except coefficient for widowed persons in urbanareas. To by married have the highest positive ef-fect on consumption expenditure level for thewhole country and in urban areas compare to asingle households. Oppositely, in rural areas thehighest effect on welfare compare to single house-holds has the households which head was wid-owed. Cohabitation in rural areas has negativeeffect on welfare, as well as to be divorced.

Conclusions

The study uses micro data from LithuanianHousehold Budget survey 2006 to examine thedeterminants of poverty in Lithuania. This pa-per presents a simple regression model used tofind out which households characteristics haveimpact on poverty in Lithuania.

Household size defined by adult equivalentunits has significant negative effect on the wel-fare status of a household. The size of the effectof household size on poverty is not the same inurban and rural areas, in urban areas it has highernegative effect.

The importance of the head of household’sage and sex was found relatively small.

The trait which most strongly and positivelyaffected the amount of household consumptionexpenditure per capita and poverty line ratio wasthe level of education held by the head of thehousehold followed by the living place of thehouseholds. The model shows that a rural house-hold have negative effect on household well-be-ing compare to urban households. The livingcounty of Lithuania have also significant impacton living standards.

The regression analysis showed that increasesin educational attainment have an important im-pact on household welfare.

It is in fact the level of education which inthis new economic system has become the decid-ing factor in terms of professional career and theamount of income earned.

Determinants of poverty in Lithuania

62

Also quite high impact of socio-economicgroup of household head was found.

References

1. Datt, Gaurav; and Jolliffe Dean. October 1999.Determinants of poverty in Egypt: 1997. Interna-tional Food Policy Research Institute, Washing-ton, D. C., FCND discussion paper, no. 75.

2. Deaton, A. 1997. Analysis of household surveys: Amicroeconometric approach to development policy.Baltimore, Md., U.S.A.: Johns Hopkins Univer-sity Press for the World Bank.

3. European social statistics. Income, poverty and so-cial excision. Eurostat, 2000.

4. Household income and expenditure. StatisticsLithuania, 2005, 2006.

5. Townsend, Ian. Research paper 04/23 Poverty:Measures and Targets. 4 March 2004.

6. Living standard and poverty. Statistics Lithuania.2005, 2006.

7. World Bank. Poverty Manual. All, JH Revision ofAugust 8, 2005.

SKURDÀ LIETUVOJE LEMIANTYS VEIKSNIAI

Algimantas MISIÛNASMykolo Romerio universitetas

Vilniaus universitetas

Graþina BINKAUSKIENËVilniaus Gedimino technikos universitetas

Lietuvos statistikos departamentas prie Lietuvos Respublikos Vyriausybës

Santrauka. Skurdo nustatymas ir ávertinimas yra svarbi problema, su kuria susiduria visos Europos Sàjungospolitikos kûrëjai. Europos Taryba 2000 m. kovà Lisabonos susitikime vienu ið svarbiausiø Europos Sàjungos uþdaviniøávardijo skurdo ir socialinës atskirties esminá sumaþinimà. 2000 m. gruodá Europos Tarybos susitikime Nicoje valstybiøvadovai pritarë nuomonei, kad geriausiø rezultatø kovojant su skurdu ir socialine atskirtimi galima pasiekti taikantatviro koordinavimo metodà. Pagrindiniai jo elementai yra bendrø tikslø Europos Sàjungai kaip visumai apibrëþimas,nacionaliniø veiksmø planø ðiems tikslams pasiekti parengimas ir periodiðka pasiektø tikslø stebësena.

Dar prieð Lietuvai integruojantis á Europos Sàjungà buvo taikomos kovos su skurdu priemonës. 2000 m. sudarytosSkurdo maþinimo Lietuvoje strategijos metmenys, kuriø pagrindu buvo parengta ir patvirtinta Skurdo maþinimostrategijos ágyvendinimo programa.

Skurstanèiais laikomi tokie asmenys, kuriø gaunamos pajamos ir kiti iðtekliai (materialiniai, kultûriniai irsocialiniai) yra tokie maþi, kad negali uþtikrinti galimybës gyventi pakankamai gerai. Pagrindinis skurdo rodiklis yrasantykinë skurdo riba, kuri sudaro 50 procentø ðalies vartojimo iðlaidø vidurkio. Bûtent ðis rodiklis ir buvo pasirinktasatliekant ðià analizæ.

Áprastai atliekamos skurdo analizës tikslas yra iðsiaiðkinti skurdo paplitimo mastà, jo kitimo tendencijas, nustatyti,kokios gyventojø grupës labiausiai kenèia nuo skurdo. Deja, tokia analizë neleidþia ávertinti skurdà lemianèiusveiksnius bei skurdo atsiradimo prieþastis. Ið ankstesniø analiziø þinoma, kad skurstantys asmenys yra þemesnioiðsilavinimo negu neskurstantys, taèiau neaiðku, ar jie gyvena þemiau skurdo ribos todël, kad jø iðsilavinimas yraþemas, ar atvirkðèiai – jø iðsilavinimas yra þemas ir todël jie skursta.

Šiame straipsnyje bandoma iðsiaiðkinti skurdà lemianèius veiksnius. Siekiant ðio tikslo naudoti Lietuvos statistikosdepartamento atlikto namø ûkiø biudþetø tyrimo duomenys. Šio tyrimo objektas yra privatus namø ûkis. Tyrimometu yra renkama informacija apie namø ûkio nariø demografines charakteristikas, socialiná-ekonominá statusà,ekonominæ veiklà, bûsto sàlygas, iðlaidas ir pajamas ir kt. Ši analizë buvo atlikta remiantis 2006 m. tyrimo duomenimis.Ið viso buvo analizuojami 7178 namø ûkiai.

Regresinë analizë yra tinkamas metodas nustatyti skurdà lemianèius veiksnius. Tokios rûðies analizë parodopriklausomo kintamojo (ðiuo atveju vartojimo iðlaidø vienam namø ûkio nariui) ir nepriklausomø kintamøjø (ávairiøveiksniø, lemianèiø skurdà) sàryðá.

Ávertinus regresijos lygties rezultatus, galima padaryti tokias iðvadas.Neigiamà átakà namø ûkiø gyvenimo sàlygoms turi namø ûkiø gyvenamoji vieta, namø ûkio dydis, iðlaikytiniø

skaièius namø ûkyje, vaikø iki 8 metø skaièius namø ûkyje. Prieðingai, teigiamai namø ûkiø vartojimo iðlaidas veikiadirbanèiøjø skaièius namø ûkyje, namø ûkio galvos iðsilavinimas, tam tikra namø ûkio galvos socialinë-ekonominëgrupë ir net ðeiminis namø ûkio galvos statusas. Namø ûkio galvos lytis namø ûkiø gyvenimo sàlygoms neturi átakos,taèiau namø ûkio galvos amþius neigiamai veikia namø ûkio gerovæ; ði átaka nëra didelë.

Algimantas Misiûnas, Graþina Binkauskienë

63_______________________

Algimantas Misiûnas – Asociate Professor, Department of Banking and Investments, Faculty of Economics andFinance Management, Mykolas Romeris University; Department of Quantitative Methods and Modelling, Faculty ofEconomics, Vilnius University.

Mykolo Romerio universiteto Ekonomikos ir finansø valdymo fakulteto Bankininkystës ir investicijø katedros,Vilniaus universiteto Ekonomikos fakulteto Kiekybiniø metodø ir modeliavimo katedros docentas, socialiniø mokslødaktaras.

Graþina Binkauskienë – Living Standard Statistics Division, Department of Statistics to the Government of theRepublic of Lithuania, also PhD student, Faculty of Business Management Department of Finance Engineering,Vilnius Gediminas Technical University.

Statistikos departamento prie Lietuvos Respublikos Vyriausybës Gyvenimo lygio statistikos skyriaus vedëjospavaduotoja ir Vilniaus Gedimino technikos universiteto Verslo vadybos fakulteto Finansø inþinerijos katedrosdoktorantë.

Determinants of poverty in Lithuania