Small Area Models for Linking Deprivation to Local Areas in Italy ( - - PowerPoint PPT Presentation

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

1

Small Area Models for Linking Deprivation to Local Areas in Italy

Gennaro PUNZO

InGrid Summer School “Reaching out to hard-to-survey groups among the poor”

HIVA-KU Leuven, Leuven - Belgium, 30 May -3 June 2016

(draft)

2

Exploring the POVERTY PATTERNS and DIFFERENTIALS across Italian NUTS3 regions for the several dimensions of life-style deprivation

AIM OF THE WORK

STEPS

• JOI NT

ANALYSI S OF MONETARY AND SUPPLEMENTARY DEPRI VATI ON ACCORDI NG TO A MULTI DI MENSI ONAL AND FUZZY APPROACH

• MANI FEST

AND LATENT DEPRI VATI ON MEASURES (BETTI ET AL., 2006)

• BORROWI NG STRENGTH ACROSS BOTH SMALL AREAS AND

TI ME: RAO – YU MODEL (1992, 1994)

• LOOKI NG

I NTO THE POTENTI AL BACKGROUND DETERMI NANTS OF THE DI FFERENT FORMS OF POVERTY

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

3

CONSIDERING DEPRIVATION IN ITS MULTIPLE DIMENSIONS...

FUZZY SET approach (Zadeh, 1965) TOTALLY FUZZY and RELATIVE method (Cheli-Lemmi, 1995)

and, in particular,

A NEW CLASS OF MONETARY AND SUPPLEMENTARY DEPRI VATI ON MEASURES TREATI NG POVERTY AS A MATTER OF DEGREE, REPLACI NG THE TRADI TI ONAL DI CHOTOMI ZATI ON POOR/ NON POOR

FUZZY MONETARY FUZZY SUPPLEMENTARY

PROPENSITY TO INCOME POVERTY PROPENSITY TO OVERALL NON- MONETARY DEPRIVATION 

Basic Life-style Housing Deterioration Housing Facilities Secondary Life-style Environmental Problems

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

4

... investigating the extent to which the several dimensions of deprivation tend to OVERLAP for households across Italian provinces

TO COMBINE THE TWO MAIN POVERTY DIMENSIONS...

LATENT DEPRI VATI ON MANI FEST DEPRI VATI ON I NCOME POVERTY NON- MONETARY DEPRI VATI ON

MANI FEST DEPRI VATI ON

TARGET VARI ABLES. . .

MANI FEST denotes a higher degree of deprivation than the LATENT one

I NTERSECTI ON between the two f uzzy sets LATENT DEPRI VATI ON UNI ON

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

5

FUZZY SET OPERATIONS

(Betti et al., 2006)

FM FS FM FS

) , max(

i i i

FS FM L 

LATENT DEPRIVATION

) , min(

i i i

FS FM M 

MANIFEST DEPRIVATION

FUZZY MANI FEST

PROPENSITY TO BOTH MONETARY AND NON-MONETARY POVERTY HOUSEHOLDS BEI NG SUBJECT TO I NCOME POVERTY AND, AT THE SAME TI ME, TO LI FE- STYLE DEPRI VATI ON

FUZZY LATENT

PROPENSITY TO EITHER MONETARY OR NON-MONETARY POVERTY HOUSEHOLDS BEI NG SUBJECT TO AT LEAST ONE OF THE TWO PREVI OUS FORMS OF DEPRI VATI ON

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

6

A METHODOLOGICAL VIEW

In order to assess the gain, in terms of efficiency, that could be achieved by borrowing strength across BOTH SMALL AREAS AND TIME...

Survey data ECHP sample (waves 1994 – 2001)

DI RECT ESTI MATES

Auxiliary variables Istat Territorial Indicators

SYNTHETI C ESTI MATES

DATA SOURCES

RAO AND YU MODEL (1992, 1994)

as ext ension of t he basic Fay–Her r iot (1979)

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

7

i i i i i

e v z x     

'

ˆ

with i = 1, 2, ... , m

Fay – Her r iot , 1979

independent and identically distributed random variables with mean 0 and variance

2 v



model on the θit’s depends on both area-specific effects (vi ) and the area-by-time specific effects (uit ) which are correlated across time for each i

Rao – Yu, 1994

it it i i it it

e u v z x      

'

ˆ

A METHODOLOGICAL VIEW

1

1 ,

  



 

it t i it

e u u

uit’s

are assumed to follow a common first order Auto-Regressive process (AR1) for each i with i = 1, 2, ... , m and t = 1, 2, ... , T

UNDER THI S MODEL, THE EBLUP COMPOSI TE ESTI MATOR I S OBTAI NED

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

8

REGRESSION COEFFICIENT ESTIMATES AND STANDARD ERRORS

Territorial Indicators (independent variables) FUZZY MANIFEST FUZZY LATENT Intercept Unemployment Rate Territorial Concentration Rate of the Resident Population

• 0.0149 (0.0518)

0.2698 (0.1222)

• 0.1007 (0.0426)

0.2297 (0.0125) 0.4606 (0.0978) –

Source: Our elaborations on ECHP data, Italian Section (1994-2001), and Istat

FUZZY MONETARY Unemployment Rate, Resident Population per 100 inhabitants, Marriage Rate FUZZY SUPPLEMENTARY Unemployment Rate, Public Hospitalization Rate, Crime Rate HEAD COUNT RATIO (traditional poverty measure) Unemployment Rate, Resident Population per 100 inhabitants, Growth Enterprises Rate, Activity Rate

OBVI OUSLY, RAO- YU MODELS HAVE ALSO BEEN ESTI MATED FOR:

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

9

I N ORDER TO EVALUATE THE PERFORMANCE OF THE ESTI MATI ON PROCESS THROUGH SMALL AREA MODELS. . .

I T ALLOWS TO TEST THE EXTENT TO WHI CH THE MODELI NG MODI FI ES THE DI RECT ESTI MATES

Direct Estimate EBLUP Composite Estimate Standard Error Direct Estimate Standard Error EBLUP Estimate

I T MEASURES THE I MPROVEMENT I N THE ACCURACY LEVEL OF THE ESTI MATES BY MODELI NG

EBLUP Estimate / Direct Estimate Mean CV Min Max FUZZY MANIFEST FUZZY LATENT 1.0347 1.0261 0.3670 0.2190 0.0910 0.3197 2.0702 1.6853 SE (EBLUP Estimate) / SE (Direct Estimate) FUZZY MANIFEST FUZZY LATENT 0.5618 0.5227 0.3720 0.3940 0.0337 0.0252 0.8827 0.9429

Source: Our elaborations on ECHP data, Italian Section (1994-2001), and Istat

Summary statistics on performance outcome measures

1 – 0.5618 = 0.4382

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

10

0.1 0.2 0.3 0.4 0.5

0.05 0.1 0.15 0.2 0.25 0.3 0.35

0.1 0.2 0.3

0.05 0.1 0.15 0.2 0.25 0.3 0.35

0.1 0.2 0.3

0.05 0.1 0.15 0.2 0.25 0.3 0.35

0.1 0.2

0.05 0.1 0.15 0.2 0.25 0.3 0.35

0.1 0.2 0.3 0.4 0.5

0.05 0.1 0.15 0.2 0.25 0.3 0.35

0.1 0.2 0.3 0.4 0.5

0.3 0.35 0.4 0.45 0.5 0.55 0.6

SOME DETERMI NANTS OF I NCOME AND LI FE- STYLE DEPRI VATI ON. . .

UNEMPLOYMENT RATE VS HCR UNEMPLOYMENT RATE VS FM UNEMPLOYMENT RATE VS FS UNEMPLOYMENT RATE VS MANI FEST UNEMPLOYMENT RATE VS LATENT ACTI VI TY RATE VS HCR

NORTH- WEST NORTH- EAST

SOUTH ISLANDS

CENTRO

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

11

Distribution of Italian NUTS3 regions by classes of poverty intensities

Head Count Ratio Fuzzy Monetary Fuzzy Supplementary Fuzzy Manifest Fuzzy Latent

< 0.05 0.05 |– 0.10 0.10 |– 0.15 0.15 |– 0.20 0.20 |– 0.25 0.25 |– 0.30 > 0.30 6.45 30.11 26.88 4.30 3.23 7.53 21.50 0.00 3.23 44.09 35.48 12.90 4.30 0.00 0.00 1.07 41.94 40.86 16.13 0.00 0.00 36.56 53.76 9.68 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 36.56 43.01 20.43

Source: Authors’ elaborations on ECHP data, Italian Section (1994-2001), and Istat

EXPLORI NG POVERTY PATTERNS AND DI FFERENTI ALS. . .

• Nearly 57% of I talian provinces shows an income poverty incidence (HCR)

between 5% and 15% while a substantial share of provinces (32%), with a poverty incidence higher than 20%, is located in the South and I slands

• More than 92% of I talian provinces shows a FM between 10% and 25%

while there isn’t any province with a FM higher than 30% or lower than 5%

• Almost the totality (98. 93%) of the I talian provinces shows a FS between

10% and 25%, denoting lower levels of territorial disparities

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

12

0.1 0.2 0.3 North-West North-East Center South Islands

EXPLORI NG POVERTY PATTERNS AND DI FFERENTI ALS. . .

0.1 0.2 0.3 0.4 0.5 North-West North-East Center South Islands

TERRI TORI AL SERI ES OF POVERTY COMPOSI TE ESTI MATES AT A NUTS3 LEVEL

HEAD COUNT RATI O (continuous line) VS FUZZY MONETARY (broken line) FUZZY MONETARY (broken line) VS FUZZY SUPPLEMENTARY (continuous line)

• The territorial series of HCR, substantially stable across northern I talian

provinces, rapidly increases as we move to the southern and insular ones

• The territorial series of FM is quite stable across northern provinces and

it slightly tends to increase as we move to the southern and insular ones

• Supplementary deprivation increases across provinces with the increasing of

the income poverty

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

13

1 2 3 4 5 North-West North-East Center South Islands 0.5 1 1.5 2 North-West North-East Center South Islands

EXPLORI NG POVERTY PATTERNS AND DI FFERENTI ALS. . .

RATI O FM/ HCR (broken line) VS FS/ HCR (continuous line) RATI O FUZZY SUPPLEMENTARY / FUZZY MONETARY

• Although the FM/ HCR territorial series appears to be relatively steady

across I talian provinces, it slightly tends to decrease with increasing HCR

TERRI TORI AL SERI ES OF POVERTY COMPOSI TE ESTI MATES AT A NUTS3 LEVEL

• I t is important to detect that, f or the same or similar HCR values, a more

severe poverty condition could be revealed f or those provinces with a higher degree of f uzzy income poverty

• The degree of f uzzy non- monetary deprivation in those provinces with a

high degree of f uzzy income poverty, i.e. , the southern and island ones, is usually lower than the corresponding level of monetary poverty

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

14

EXPLORI NG POVERTY PATTERNS AND DI FFERENTI ALS. . .

FUZZY MANI FEST (continuous line) VS FUZZY LATENT (broken line)

0.1 0.2 0.3 0.4 North-West North-East Center South Islands

• Although both the MAN and LAT territorial series show an upward

trend f or the poorer southern provinces, the MAN always states a degree of poverty largely higher than the LATENT one

TERRI TORI AL SERI ES OF POVERTY COMPOSI TE ESTI MATES AT A NUTS3 LEVEL

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

15

Distribution of Italian NUTS3 regions by classes of poverty intensities

Head Count Ratio Fuzzy Monetary Fuzzy Supplementary Fuzzy Manifest Fuzzy Latent

< 0.05 0.05 |– 0.10 0.10 |– 0.15 0.15 |– 0.20 0.20 |– 0.25 0.25 |– 0.30 > 0.30 6.45 30.11 26.88 4.30 3.23 7.53 21.50 0.00 3.23 44.09 35.48 12.90 4.30 0.00 0.00 1.07 41.94 40.86 16.13 0.00 0.00 36.56 53.76 9.68 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 36.56 43.01 20.43

Source: Authors’ elaborations on ECHP data, Italian Section (1994-2001), and Istat

EXPLORI NG POVERTY PATTERNS AND DI FFERENTI ALS. . .

• Any I talian province shows a MAN estimate higher than 15% so

as any province denotes a LAT estimate lower than 20%

• Although at

a dif f erent level, the MAN and LAT estimates show an adequate co- graduation degree (t =

• 0. 50)

across I talian provinces

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

16

EVALUATI NG THE DEGREE OF OVERLAP. . .

RATI O FUZZY MANI FEST / FUZZY LATENT

• The

degree

• f

OVERLAP between income and lif e- style deprivation at the level of household is higher in the poorer southern and insular provinces than the richer northern ones

TERRI TORI AL SERI ES OF POVERTY COMPOSI TE ESTI MATES AT A NUTS3 LEVEL

0.05 0.1 0.15 0.2 0.25 0.3 0.35 North-West North-East Center South Islands

     1 ... LAT MAN

NO- OVERLAP SI MULTANEOUSLY DEPRI VED

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

17

CONCLUDING REMARKS

Remar kable GAI NS I N EFFI CI ENCY of t he pover t y est imat es AND t he high level of STATI STI CAL SI GNI FI CANCE

• f some t er r it or ial indicat or s highlight t he model adequacy

RAO AND YU MODEL – EBLUP ESTI MATOR

• Even though at

a dif f erent level, both the monetary and the living conditions approach broadly conf irm the distinctive territorial socio- economic gap between the North and South of I taly

• Although

the deprivation patterns are basically unchanged, the territorial distances between the northern provinces and the southern

• nes appear to be more marked by the conventional approach than the

f uzzy income one

• Supplementary deprivation increases across provinces with the increasing
• f the income poverty even though the living conditions seem to be more

severe than the income deprivation across northern provinces

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies

18

CONCLUDING REMARKS AND DEVELOPMENTS

EBLUP Estimators ar e var iable-specif ic since t hey may depend on t he par t icular pover t y measur e consider ed in t he small ar ea models

Theref ore, in order t o assess st ill f urt her t he RELATI VE PERFORMANCE

• f DI RECT, SYNTHETI C and COMPOSI TE EBLUP est imat ors associat ed t o

Rao and Yu models adopt ed in t his work...

• SI MULATI ON STUDY
• A set of QUALI TY I NDI CATORS, i.e., AARB, AARE, AEFF, ARMSE, ...

Subj ect ive Pover t y f ocused on t he abilit y t o “make ends meet ” Fr om ECHP t o EU-SI LC

FI NALLY. . .

2nd J une 2016 Universit y of Naples “Part henope” (I t aly) Depart ment of Management and Quant it at ive St udies