Comparing poverty estimates using income, expenditure and material deprivation. Paola Serafino and Richard Tonkin

Transcription

1 14 EU-SILC International Conference Conference, Lisbon, October 14 - Session 1 - Comparing poverty estimates using income, expenditure and material deprivation Paola Serafino and Richard Tonkin (UK Office for National Statistics [ONS]) This paper was prepared as part of "Net-SILC2", an international research project funded by Eurostat and coordinated by CEPS/INSTEAD (Luxembourg). It is a draft version subject to revision. It should not be quoted or disseminated without the written consent of the author(s).

2

3 Comparing poverty estimates using income, expenditure and material deprivation Authors: Paola Serafino and Richard Tonkin (1) Abstract: The Europe social inclusion target will be measured according to work attachment, income and material deprivation indicators using the EU Statistics on Income and Living Conditions (EU-SILC). However, there has been increasing interest in recent years in whether expenditure and consumption provide more appropriate measures of standard of living than income. This Net-SILC2 work package therefore aims to compare people s exposure to poverty using four different measures: income, expenditure, material deprivation and low work intensity. However, no single data source provides joint information on all these variables. Therefore, the analysis described in this paper uses the results of statistically matching expenditure from the Household Budget Survey (HBS) with income and material deprivation contained within EU Statistics on Income and Living Conditions (SILC) using data for the UK, Germany and Belgium. These matched datasets were used to analyse the overlap between income and expenditure poverty and material deprivation, as well as the relationship between income and expenditure poverty and other measures of social exclusion. (1) Richard Tonkin and Paola Serafino are from the UK Office for National Statistics (ONS). This work has been supported by the second Network for the analysis of EU-SILC (Net-SILC2), funded by Eurostat. The European Commission and ONS bear no responsibility for the analyses and conclusions, which are solely those of the authors. address for correspondence: paola.serafino@ons.gsi.gov.uk. The authors would like to thank David Gordon, Tony Atkinson and Eric Marlier for their helpful comments and discussions. 1

4 1. Introduction Most policy initiatives aimed at improving living standards tend to measure poverty relatively within the society, using income as a yardstick. However, there is an argument that income isn t sufficient as a sole measure of poverty, particularly if one considers poverty in terms of achieved standards of living 1. It is the consumption of goods and services, along with other inputs such as time that ultimately satisfies a household s wants. Because of this, it is arguably a more important determinant of economic well-being than income alone. Indeed, Brewer & O Dea (12) and others (see Noll, 7, for a review) argue that it is preferable to consider the distribution of consumption rather than income on both theoretical and pragmatic grounds. On a theoretical ground, income can be subject to fluctuations, due to such events as short-term unemployment. However, these fluctuations in income are not likely to be matched by corresponding downturns in living standards, as households are typically able to smooth consumption by drawing on savings or help from family members. This finding leads to Friedman s permanent income hypothesis, which suggests that decisions made by consumers are based on long-term income expectations rather than their current income. This view is supported in a number of studies (e.g. Cutler & Katz, 1991, and Jorgenson & Slesnick, 1987) which find stronger relationships between consumption and subjective well-being than between income and subjective well-being measures. Beyond these conceptual arguments, there is also the practical consideration that evidence from a range of countries suggests a general tendency for income to be under-reported by households with low levels of resources, whilst reporting of expenditure by this group is relatively accurate (e.g. Meyer & Sullivan, 11 and Brewer & O Dea, 12), though other evidence suggests that expenditure of higher income households may be under-reported (Sabelhaus, et al., 11). In economic and social research, data on household expenditure are typically used as a proxy for consumption. These data are often collected through the use of diary studies. However, it should be noted that expenditure is an imperfect measure of consumption as the amount spent by a household in a given month may differ from consumption, due to households making use of goods purchased previously or the purchase of consumer durables. In addition, consumption also includes inter-household in-kind transfers of gifts and services and social transfers in kind. However, these aspects of consumption are generally excluded from data due to the challenges of collecting this type of information. Overall the evidence indicates that while income can be a good proxy for living standards, it is better when supplemented with a wider range of measures such as expenditure. This is consistent with the recommendations of the Report by the Commission on the Measurement of Economic Performance and Social Progress (Stiglitz, Sen, & Fitoussi, 9) as well as the OECD Framework for Statistics on the Distribution of Household Income, Consumption and Wealth (13). This Net-SILC2 work package therefore aims to compare people s exposure to poverty using four different measures: income, expenditure, material deprivation and low work intensity, across countries of the EU. However, there is no data source which provides joint information 1 As well as considering poverty in terms of an individual s standard of living, other approaches are possible, such as considering poverty in terms of a right to a minimum level of resources (see Atkinson et al. (2) for a discussion). 2

5 on all of these variables for households or individuals. Therefore, the first stage of this project involved statistically matching expenditure from the Household Budget Survey (HBS) with income and material deprivation contained within EU Statistics on Income and Living Conditions (SILC). Preliminary work was carried out to develop the methodology using 5 UK data (see Webber & Tonkin, 13). This paper builds on that work by first presenting the results of statistical matching of HBS and EU-SILC data for a number of countries using data, before going on to use that data to carry out joint analysis of income and expenditure based poverty and other measures of disadvantage, including severe material deprivation. The current version of this paper presents results for the UK, Germany and Belgium and work is underway to extend the matching and analysis to a number of other EU countries. The selection of countries has been constrained by both restrictions on access to HBS microdata and the suitability of the two data sources for statistical matching. 2. Statistical matching 2.1 Overview of statistical matching Statistical (or synthetic) matching is a broad term used to describe the fusing of two datasets. In this context, the datasets are of households sampled from the same population. The usual approach is to define one data set as the recipient, in this case EU-SILC, and one as the donor, HBS. The recipient data contains a variable Y, in this case material deprivation, which is not found in the donor, while variable Z, expenditure, is only contained within the donor. The aim is to use information contained within the set of variables common to both datasets, X, to link records from the donor to the recipient. Therefore, expenditure is linked to EU-SILC, which contains information on income, material deprivation and work intensity. Recipient dataset (EU-SILC) Matched dataset Y X Donor dataset (HBS) X Z Y X Z 3

6 2.2 Reconciliation of the data sources In order for statistical matching to be a success, it is vital that steps are taken to ensure the donor and recipient datasets, the variables and their distributions are comparable. D Orazio, Di Zio, & Scanu (6 pg 164) outline the following eight steps for achieving this: Harmonization of the definition of units. Harmonization of reference periods. Completion of population. Harmonization of variables. Harmonization of classifications. Adjustment for measurement errors (accuracy). Adjustment for missing data. Derivation of variables. Before carrying out statistical matching it is necessary to ensure that the key concepts are defined in a comparable way in the donor and recipient, in this case, the definitions of household, household reference person, population and income reference period Household The concept of a household is similarly defined for both HBS and EU-SILC. This definition states that a household is constituted by a person or people living together in the same dwelling who share meals or joint provision of living conditions Household reference person In HBS the household reference person (HRP) is clearly defined and identified. The HRP is the householder who: owns the household accommodation, or is legally responsible for the rent of the accommodation, or has the household accommodation as an emolument or perquisite, or has the household accommodation by virtue of some relationship to the owner who is not a member of the household. If there are joint householders the household reference person will be the one with the higher income. If the income is the same, then the eldest householder is selected. In EU-SILC, there is no household reference person as such. However, there are identifiers for up to two people responsible for the household accommodation. These identifiers are defined in a similar way to the HRP on HBS, except that in the case of joint householders, the default is to report the oldest householder, with no consideration of income. Since 1/2 the concept of household reference person (HRP) has been adopted on all UK Government sponsored surveys. Therefore, the definition of HRP is the same on EU-SILC and HBS for the UK. This is not necessarily the case for the other countries in the analysis. 4

7 2.2.3 Population and Sampling Frame In all countries studied, both sources cover the same population (private households, excluding collective establishments). In the UK, both sources also use the same sampling frame (the Postcode Address File a list of addresses provided by the UK Post Office). In Germany, the sampling frames are different. EU-SILC uses a random sample of households who have responded to the German microcensus and have agreed to participate in further voluntary surveys. The sample for the German Household Budget Survey are largely selected from respondents to the sample survey of household income and expenditure (EVS). The sampling frame for EU-SILC and the HBS in Belgium is the Central Population Register. This Register includes all private households and their current members residing in the territory Reference Period EU-SILC in the UK measures current income. Therefore, for the UK in, both the EU-SILC dataset and the HBS dataset measure current annual income in For Germany and Belgium, the income reference period for SILC and HBS data is the previous calendar year, so 9 in the case of data. 2.3 Harmonization of variables Annex 1 contains the full list of variables common to both EU-SILC and HBS in. The original statistical matching methodology set out in Webber & Tonkin (13) was developed using the 5 HBS, which included a number of variables related to ownership of material goods that were dropped from the survey. In the UK, these data were still collected on the Living Costs and Food Survey (LCF), the survey that is used to derive the HBS variables. This meant that it was possible to merge these variables onto the HBS for the UK, and allowed them to be included in the matching process for this country. This was not possible for the other countries examined. Annex 2 contains details of the variables taken from the LCF for the UK and their comparable variables in EU-SILC. The variables common to both datasets needed to be harmonized across the two sources in order to be used for the matching. This involved recoding of variables to the stage where they have the same degree of detail. The table in Annex 3 shows the codification of these derived variables. The HBS variable that defines activity status, for instance, is more detailed than the corresponding EU-SILC variable. The detail in HBS therefore needed to be sacrificed to ensure that it is comparable with EU-SILC. This highlights a constraining factor in statistical matching that detailed information on one survey is lost unless the corresponding variable on the other data set is available at the same level. Once the variables have been harmonized a check for missing information was performed because some of the statistical matching methods used rely on regressions. If a variable has missing information in one case, that whole case is omitted from the regression, thereby losing potentially valuable information from the other variables. Where missing information would have resulted in the loss of too many cases, variables were excluded from further analysis. 5

8 2.4 Choosing the matching variables The variables selected for matching must fulfil two criteria. First, there must be similarity in the distributions of the variables across the two surveys. Second, the variables must be significant in explaining variations in the target variables in this case expenditure and material deprivation Coherence of distributions The literature highlights two main methods for calculating the degree to which distributions of variables are similar across data sets. The first is a simple comparison of the weighted frequency distributions of the derived variables in the two datasets. The second is to use a measure such as the Hellinger Distance (HD). The HD is convenient because it provides a single number as a measure for the similarity in distribution of two variables. There is no fixed rule regarding what degree of similarity is suitable for statistical matching purposes, though Leulescu & Agafitei (13) suggest that a HD of over 5% should raise concerns about the similarities in distributions. The equation used to derive the HD is: K HD (V, V ) = 1 2 ( n Oi n Pi ) N O N P i=1 Variable V is in the donor data set, V in the recipient, K is the total number of cells in the contingency table, n Oi is the frequency of cell i in the original data O, n Pi is the frequency of cell i in the recipient and N is the total size of the specific sources. Table 1 shows the Hellinger Distances for the common variables for each of the countries in the analysis. Missing values generally reflect that the variable(s) required were not available on one of the datasets. Where the HD was found to exceed 5% for the potential matching variables, various options were explored. Two outcomes could be coded to a single one to overcome large discrepancies in the original proportions of the outcomes. This can reduce the HD thereby ensuring that it is suitable as a potential matching variable. However, by limiting the possible outcome responses in the variable reduces its variation, thereby making it potentially less likely to be useful in explaining variations in material deprivation or expenditure. For example, as the HD for DV_AGE2 was relatively high for Belgium (9.1) an alternative age variable including fewer categories was created (DV_AGE3). Consideration was also given to recoding outcomes with a high divergence as missing observations to be excluded from the analysis. Although this can reduce the HD to an acceptable level, it can remove an unacceptable number of observations. 2 6

9 Table 1: Hellinger distances (HD) of EU-SILC and HBS variables, Variable HD (%) UK GERMANY BELGIUM DV_SEX DV_AGE DV_AGE DV_REGION DV_URBAN DV_HHSIZE DV_HHTYPE (Household type) DV_DWELL DV_ROOMS DV_TENURE DV_MARSTA (Marital status) DV_CONUNI (Consensual union) DV_MAXEDU (Educational attainment) DV_LABOUR DV_ACTSTAT (Activity status) DV_ACTSTAT2 (Activity status) DV_OCC DV_CAR DV_TV DV_PC. - - DV_WASH DV_PHONE INC_BAND Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The Hellinger distance provides a measure of the similarity in distributions of two variables, in this case the variables derived in the HBS and EU-SILC. A Hellinger distance of under 5 indicates a comparable distribution of the variables being compared. Having explored the potential recoding options described, where the HD remained in excess of 5%, the variables were generally dropped. An exception to this was the variables DV_OCC and DV_DWELL for the UK; these were retained despite having an HD marginally in excess of 5% because the weighted frequencies were sufficiently similar to warrant inclusion. In contrast, for the UK data, marital status, DV_MARSTA, had an HD of 6.9%, and comparison of the weighted frequency distributions for this variable revealed large differences in the proportion of people identified as being married. This was due to cohabitation being included in the married category for HBS but not for EU-SILC. As a result, this variable was dropped. This highlights the importance for effective statistical matching of ensuring that definitions for common variables are harmonised across data sources. Table 2 shows the variables retained for each country. 7

10 Table 2: Variables retained for matching process, VARIABLE UK GERMANY BELGIUM DV_SEX DV_AGE2 DV_AGE3 DV_REGION DV_URBAN DV_HHSIZE DV_HHTYPE (Household type) DV_DWELL DV_ROOMS DV_TENURE DV_MARSTA (Marital status) DV_CONUNI (Consensual union) DV_LABOUR DV_MAXEDU DV_ACTSTAT (Activity status) DV_ACTSTAT2 (Activity status) DV_OCC DV_CAR DV_TV DV_PC DV_WASH DV_PHONE INC_BAND Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS Explanatory power of the variables D Orazio et al (6) identifies the following method for choosing the matching variables from the set of common variables: 1. Let ψ A consist of all the common variables such that ψ A is independent of Y given the other common variables in the recipient data set. 2. Let ψ B consist of all the common variables such that ψ B is independent of Z given the other common variables in the donor data set. 3. Let ψ=ψ A ψ B; then the other common variables define X, the matching variables. Therefore, the common variables which were used for matching were those that are statistically significant in explaining variations in both expenditure and material deprivation. Material deprivation was defined as a binomial variable, taking a value of 1 if the respondent was materially deprived and otherwise. A logistic regression was fitted to model deprivation using the variables shown in Table 2 for each country. Next, an expenditure model was estimated on HBS data. As expenditure is highly positively skewed, the stepwise regression model for each country was estimated on the logarithm of expenditure, using the same variables as before. 8

11 As stated above, the variables that should be selected for matching are those which are significant in explaining material deprivation and expenditure. Therefore, the final matching variables for each country are shown in Table 3. Table 3: Final matching variables, VARIABLE UK GERMANY BELGIUM DV_SEX DV_AGE2 DV_AGE3 DV_REGION DV_URBAN DV_HHSIZE DV_HHTYPE (Household type) DV_DWELL DV_ROOMS DV_TENURE DV_MARSTA (Marital status) DV_CONUNI (Consensual union) DV_LABOUR DV_MAXEDU DV_ACTSTAT (Activity status) DV_ACTSTAT2 (Activity status) DV_OCC DV_CAR DV_TV DV_PC DV_WASH DV_PHONE INC_BAND Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. 2.5 Matching methods Three different matching methods were used in this analysis, covering the three broad categories of approaches typically used in statistical matching: Non-parametric methods Parametric methods Mixed methods The hotdeck method is a non-parametric approach. The procedure finds records in the donor file and matches them with records in the recipient file, based on a distance function. This results in actual observed values, for expenditure in this case, being imputed onto EU-SILC. A disadvantage of this procedure, and especially relevant in this scenario, is that the multiple usage of donors is necessary as the donor dataset, HBS, is smaller than the recipient, EU-SILC. This can increase the risk that the distribution of the imputed variable does not reflect the original one. The second (parametric) approach involves imputing predicted values obtained from a regression model. The reliability of this method is very much dependent on the accuracy of the model. In addition, regression towards the mean can be a potential problem with this approach. 9

12 Mixed methods, as the name implies, involves a combination of parametric and non-parametric techniques. A model is first fitted to the data to estimate an intermediate value of the variable to be matched. Then a distance function is used to locate a range of possible observations from the donor set which most closely resembles the intermediate value, with a value for imputation selected from that set. In the method used, this process was performed multiple times, producing multiple imputed datasets. This builds in some allowance for uncertainty in the model. Analysis was carried out on each imputed dataset, before the results were averaged across the imputed datasets to produce one overall set of estimates. 3. Results of statistical matching Testing the validity of matching procedures involves comparing the distributions of the matched variables against observed expenditure in the HBS. This was done in three ways: By comparing mean expenditure by equivalised expenditure decile to analyse the consistency of the overall expenditure distribution for each method. By comparing the consistency of mean expenditure by variables used in the statistical matching for observed and imputed expenditure. By comparing the relationship between expenditure and variables in both datasets but not included in the model. The following section provides results of some of the main comparisons that were carried out, for all countries studied. Further details of the comparisons not reported here due to space constraints are available on request. 3.1 Comparison of mean expenditure by expenditure deciles EU-SILC imputed versus HBS observed, Figure 1 provides an indication of the performance of the different matching methods across the expenditure distribution. All three methods appear to be relatively effective in replicating mean expenditure by expenditure deciles. Table 4 compares the mean total expenditure in the HBS with each of the matched datasets. For the UK and Belgium mean expenditure most closely matches the HBS value for the mixed methods approach. For Germany the Hotdeck approach provides the closest match though in all cases the mean values are fairly close to those seen in the original HBS. This further supports the view that all three methods are providing effective matches. Table 4: Mean expenditure for HBS and each of the matching methods, ( per annum) Country HBS Mixed methods Hotdeck Parametric UK 25,56 25,481 25,74 25,34 Germany 29,33 28,966 29,13 28,965 Belgium 34,46 34,5 33,786 33,57 Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The table compares the mean expenditure in the original HBS with mean expenditure in each of the EU-SILC matched datasets.

13 Milliers Milliers Milliers Figure 1: Mean expenditure by equivalised household expenditure decile for HBS and different matching methods, ( per annum) 7 6 UK Germany Belgium HBS Mixed Hotdeck Parametric Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figures compare results from the original HBS with the EU-SILC datasets matched using each of the three methods described. It shows for each dataset, the mean expenditure in per annum of households in each of the expenditure deciles. The deciles are calculated using expenditure which has been equivalised to take into account the household composition. The quality of the matching process is determined by how closely the distribution in the matched datasets corresponds to the original HBS, shown by the red bar. 11

14 Milliers Milliers Milliers 3.2 Comparison of expenditure by matching variables - EU-SILC imputed versus HBS observed Figure 2: Mean total household expenditure by unequivalised income band for HBS and matching methods, ( per annum) United Kingdom Under 5, 5, - 9,999, - 14,999 15, - 19,999, - 29,999 3, - 39,999 4, - 49,999 5, or more 6 5 Germany 4 3 Under 5, 5, - 9,999, - 14,999 15, - 19,999, - 29,999 3, - 39,999 4, - 49,999 5, or more 7 6 Belgium Under 5, 5, - 9,999, - 14,999 15, - 19,999, - 29,999 3, - 39,999 4, - 49,999 5, or more HBS Mixed Hotdeck Parametric Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: These figures compare the mean expenditure in per annum of households banded according to the total household income from the original HBS with the EU-SILC datasets matched using each of the three methods described. 12

15 Figure 2 shows the distribution of actual total household expenditure in the HBS, and expenditure derived from the matching methods across the income distribution. All three methods appear to perform well in general. At the low end of the income distribution we see the expected expenditure tick higher average expenditure for the bottom income group than households in the second income group. From Figure 2, while the hotdeck method appears to perform best for Belgium, for the remaining countries none of the methods appears consistently better than the others at matching across the income distribution. 3.3 Comparison of expenditure by matching variables observed versus imputed HBS Another way of assessing the quality of the matching processes is to artificially remove expenditure from a random selection of half the HBS sample and then impute expenditure back on using each of the three methods. Figure 3 shows the distribution of mean expenditure by equivalised expenditure decile using this approach. Again, all three methods appear relatively effective at replicating the expenditure distribution in the HBS, though with some underestimation of the higher deciles for the German data. Overall the mixed methods approach provides the closest match across the distribution as a whole for all three countries. 13

16 Milliers Milliers Milliers Figure 3: Mean household expenditure by equivalised expenditure decile for HBS observed and HBS imputed, ( per annum) United Kingdom Germany Belgium HBS Mixed Hotdeck Parametric Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: In these figures each method was used to impute expenditure figures onto a random sample of half the HBS respondents, that had had their expenditure set to. It shows for each dataset, the mean expenditure in per annum of households in each of the expenditure deciles. The deciles are calculated using expenditure which has been equivalised to take into account the household composition. The quality of the matching process is determined by how closely the distribution in the matched datasets corresponds to the original HBS, shown by the red bar. 14

17 Milliers Milliers 3.4 Comparison of expenditure by variables not used in statistical matching Figure 4: Mean household expenditure by household type for HBS and matching methods, ( per annum) 45 4 United Kingdom Adult 2 Adults >2 adults 1 Adult with DCH 2 Adults with DCH >2 adults with DCH 6 Germany Adult 2 Adults >2 adults 1 Adult with DCH HBS Mixed Parametric Hotdeck 2 Adults with DCH >2 adults with DCH Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: These figures compare results from the original HBS with the EU-SILC datasets matched using each of the three methods described. It shows for each dataset, the mean expenditure in per annum of households based on the household composition. DCH stands for dependent children. The quality of the matching process is determined by how closely the distribution in the matched datasets corresponds to the original HBS, shown by the red bar. Figure 4 shows the relative performances of the matching methods at estimating expenditure across a variable not used in the matching process household type. Results are only presented for the UK and Germany as household type was a matching variable for Belgium. All three methods perform reasonably well, particularly for the German data. For the UK there is some over/under-estimation of expenditure for certain types of household for all methods. In particular, expenditure appears to be overestimated for single adult households, but underestimated for households with more than two adults. 15

18 4. Income and expenditure poverty analysis The following section presents joint analysis of income and expenditure based poverty and severe material deprivation across various EU countries. The results of the statistical matching, presented in the previous section, revealed that all three approaches to statistical matching were broadly effective, with the results for the mixed and hotdeck methods marginally better than those for the parametric approach. Based on this, it was decided to use the outputs of the mixed methods approach for the analysis section of this paper. For the purpose of this analysis, income poverty is defined as having an equivalised household income below 6% of the national equivalised median income. This is in line with the definition used in the At Risk of Poverty or Social Exclusion (AROPE) indicator which is used to monitor progress towards the Europe headline target. The main expenditure poverty measure used is defined in comparable terms: equivalised household expenditure less than 6% of the equivalised median. Where the analysis uses other thresholds for expenditure poverty, these are clearly identified. Individuals are classed as being severely materially deprived if they have an enforced lack of at least four out of a list of nine material deprivation items 2. The work intensity of a household is defined as the ratio between the total number of months that all working-age household members (aged excluding students) have worked and the total number of months those household members could have worked. Someone is defined as low work intensity if they are living in a household with work intensity less than Headline poverty indicators Figure 5 shows how the estimates of income and expenditure poverty compare between the HBS and the matched EU-SILC datasets. In general there is a relatively close correspondence between estimates of income poverty in the two datasets for all three countries. The country with the largest divergence is the UK, where the HBS produces slightly higher estimates of income poverty than EU-SILC (18.3% and 16.1% respectively), while the reverse is true for expenditure poverty (15.% on HBS and 18.3% on EU-SILC). 2 Currently these 9 items are: Arrears on mortgage or rent payments, utility bills, hire purchase instalments or other loan payments; capacity to afford paying for one week s annual holiday away from home; capacity to afford a meal with meat, chicken, fish (or vegetarian equivalent) every second day; capacity to face unexpected financial expenses; household cannot afford a telephone (including mobile phone); household cannot afford a colour TV; household cannot afford a washing machine; household cannot afford a car; ability of household to pay for keeping its home adequately warm. 16

19 EU-SILC EU-SILC Figure 5: Income and expenditure poverty in matched EU-SILC and HBS, (% population) 3 a) Income poverty 3 b) Expenditure poverty UK 15 DE BE UK 15 DE BE HBS HBS Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figures compare the percentage of the total population of each country in a) income poverty and b) expenditure poverty on the HBS with those on the matched EU-SILC dataset. The dotted line indicates where the points should lie if both surveys produced identical estimates for these measures. Figure 6 compares estimates of income and expenditure poverty in the HBS and matched EU- SILC datasets, using alternative thresholds for expenditure poverty, including 5% and 4% of median equivalised expenditure, as well as 6% of median equivalised income. In the UK and Germany, expenditure poverty is higher in EU-SILC than in the HBS for all the thresholds examined, indicating that the difference observed in Figure 5 is not simply an artefact of the threshold used. For Belgium, the expenditure poverty rates were very close on EU-SILC and HBS for all three expenditure thresholds used. Expenditure poverty in Belgium was slightly higher on the HBS when 6% of median equivalised income was used as the threshold. One notable feature of these figures is that expenditure poverty is considerably higher in the UK when the threshold is based on income than expenditure. This reflects median income being considerably higher than median expenditure in the UK HBS data and therefore EU-SILC. This difference may be partly due to mortgage interest payments not being reflected within the main expenditure variable in HBS (HE). However, this does not explain why the same difference is not present in the German and Belgian data. 17

20 Figure 6: Income and expenditure poverty in matched EU-SILC and HBS using different thresholds for expenditure poverty, UK (% population) 4 3 United Kingdom SILC HBS Income poor Expenditure poor Expenditure poor Expenditure poor Expenditure poor (6% exp median) (5% exp median) (4% exp median) (6% income median) Germany SILC HBS 15 5 Income poor Expenditure poor (6% median) Expenditure poor (5% median) Expenditure poor (4% median) Expenditure poor (6% income median) Belgium SILC HBS Income poor Expenditure poor (6% median) Expenditure poor (5% median) Expenditure poor Expenditure poor (4% median) (6% income median) Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figures compare estimates of income and expenditure poverty on the HBS with the EU-SILC dataset matched using the mixed methods approach using different thresholds for expenditure poverty; the threshold for expenditure poverty is shown in each case in brackets. 18

21 4.2 Overlap of income and expenditure poverty and material deprivation Figure 7 shows the percentage of the population experiencing poverty on one or more of the measures and the overlap between them. In this figure, deprived refers to those experiencing severe material deprivation. These Venn diagrams show that the degree of overlap between the three measures varies substantially across the countries examined, with the difference between the UK and Germany particularly prominent. In the UK, 31% of people were either in income poverty or expenditure poverty or were materially deprived, while 8% were in poverty on two or more of the measures and 1% were in poverty on all three. In Germany, though the proportion of people who were in poverty on at least one of the three measures was lower, at %, the degree of overlap between the measures was higher: 9% of people were in poverty on two or more and 2% were in poverty on all 3. This figure also shows a stronger link between income poverty and severe material deprivation than expenditure poverty and severe material deprivation. For the UK, of the 5% of the population that were severely materially deprived, while 4% were income poor only 26% were expenditure poor. In Belgium, over half of the 3.8% that were deprived were income poor while only 36% were expenditure poor. Similarly in Germany, which had the largest level of overlap between the measures, almost 6% of the 4.6% that were deprived were income poor while just over 4% were expenditure poor. 19

22 Figure 7: Breakdown of population by poverty status, (% population) a) UK b) Germany c) Belgium Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figures show the percentages of the population experiencing each form of poverty and severe material deprivation for each of the countries and how these overlap with one another. In this figure deprived refers to severe material deprivation.

23 Looking more closely at the degree of overlap between income and expenditure poverty, figure 8, shows again how this varies across the three countries examined. In the UK, only 37% of those in income poverty were also expenditure poor, whereas the overlap between the two measures was higher in Belgium (47%) and higher still in Germany (55%). Figure 8: Percentage of income poor individuals experiencing expenditure poverty, (%) % 9% 8% 7% 6% Income and expenditure 5% poverty 4% Income poverty only 3% % % % UK Germany Belgium Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figure shows the relative proportion of income poor individuals that also experience expenditure poverty. A similar pattern is observed when looking at the proportion of expenditure poor individuals who were also income poor (Figure 9). In the UK, just less than a third (32%) were also income poor, but the overlap between these groups was substantially higher in both Belgium (45%) and Germany (64%). Figure 9: Percentage of expenditure poor individuals experiencing income poverty, (%) % 9% 8% 7% 6% 5% 4% 3% % % % UK Germany Belgium Income and expenditure poverty Expenditure poverty only Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figure shows the relative proportion of expenditure poor individuals that also experience income poverty. 21

24 Figure : Percentage of severely materially deprived individuals experiencing income and expenditure poverty, (%) % 9% 8% 7% 6% 5% 4% 3% % % % UK Germany Belgium Income and expenditure poverty and material deprivation Expenditure poverty and material deprivation Income poverty and material deprivation Material deprivation only Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figure shows the relative proportion of severely materially deprived individuals that also experience income and/or expenditure poverty. In all 3 countries, there was a greater overlap between severe material deprivation and income poverty than severe material deprivation and expenditure poverty, with a higher proportion of those who were severely materially deprived also being in income poverty (Figure ). If expenditure does provide a better measure of living standards than income, it might be expected that the relationship between expenditure poverty and measures such as material deprivation would be stronger than that between income poverty and such measures. To begin to examine this point, Figure 11 shows for each poverty measure, the percentage of those in poverty and not in poverty that are experiencing severe material deprivation. For all three countries, there is a stronger relationship between income poverty and severe material deprivation than expenditure poverty and severe material deprivation. This is particularly the case for the UK, where there is very little difference between the severe material deprivation rate for expenditure poor and nonexpenditure poor individuals. The difference is considerably larger in Germany and Belgium, though in both countries the severe material deprivation rate is higher for those who are income poor than those who are expenditure poor. 22

25 Severe material deprivation Figure 11: Severe material deprivation by poverty status, (%) UK GERMANY BELGIUM Income-poor Non income-poor Expenditure-poor Non expenditure-poor Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figure shows the percentage of those in income and expenditure poverty and those not in income or expenditure poverty that are severely materially deprived; for example, for the UK the dark blue bar shows that just over 12% of the income poor population in the UK are severely materially deprived. Figure 12 shows the percentage of those with different experiences of poverty that are unable to afford the key items that are used to measure deprivation across the EU. The patterns seen in this figure are similar to those seen for deprivation as a whole (Figure 11). As with severe material deprivation overall, across all the key items, there is a stronger relationship between inability to afford most of the items and income poverty than there is with expenditure poverty. Similarly, there appears to be a stronger relationship between expenditure poverty and inability to afford these items in Germany and Belgium compared with the UK. 23

26 Figure 12: Population unable to afford key deprivation items by poverty status, (% population) UK Heating home Mortgage, rent or bills Holiday Protein Unexpected expenses Germany Car Washing machine TV Phone Heating home Mortgage, rent or bills Holiday Protein Unexpected expenses Car Washing machine TV Phone 7 Belgium Heating home Mortgage, rent or bills Holiday Protein Unexpected expenses Car Washing machine Income poor Non income poor Expenditure poor Non expenditure poor TV Phone Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: These figures compare the percentage of those in income and expenditure poverty with those not in income or expenditure poverty that are unable to afford the 9 items used to measure severe material deprivation; for example, for the UK the initial dark blue bar shows that just over % of the income poor population in the UK cannot afford to keep their homes adequately warm. 24

27 4.3 Housing related deprivation Figure 13: Population experiencing additional poor living conditions by poverty status, (%) % 35 UK Damp Private bath Private toilet Dark Noisy Pollution Crime % Germany Damp Private bath Private toilet Dark Noisy Pollution Crime % Belgium Damp Private bath Private toilet Dark Noisy Pollution Crime Income poor Non income poor Expenditure poor Non expenditure poor Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: These figures compare the percentage of those in income and expenditure poverty with those not in income or expenditure poverty that experience poor housing conditions; for example, for the UK the initial dark blue bar shows that just over 3% of the income poor population in the UK live in a house that is damp or has a leaking roof. 25

28 Low work intensity In addition to the items that are used in the main material deprivation measure, EU-SILC also includes a number of variables that are indicative of poor housing conditions. These include living in a home that is subject to damp or a leaking roof, that is dark or excessively noisy, that is in an area that suffers environmental problems or high levels of crime, or without sole use of bathing facilities or an indoor flushing toilet. Figure 13 shows the relationship of these variables with income and expenditure poverty. In Germany and Belgium, the relationship between most of these measures of poor living conditions and poverty appears to be equally strong for both income and expenditure poverty. By contrast, with the exception of living in a home that is subject to damp or a leaking roof, and high levels of crime in the local area, there is no evidence of a relationship between income poverty and these measures in the UK, and no evidence of a relationship between expenditure poverty and any of these indicators. This suggests that, in the UK, issues around properties that are dark or excessively noisy or in an area suffering from environmental problems are not directly related to relative low income. 4.4 Low work intensity The third component of the EU at-risk-of-poverty or social exclusion target is low work intensity. An individual is defined as having low work intensity status if they live in a household where the working age adults worked less than % of the time that they could have worked in the reference period. Figure 14: Low work intensity by poverty status, (%) UK GERMANY BELGIUM Income-poor Non income-poor Expenditure-poor Non expenditure-poor Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figure shows the percentage of those in income and expenditure poverty and those not in income or expenditure poverty that have low work intensity staus; for example, for the UK the dark blue bar shows that approximately 33% of the income poor population in the UK have low work intensity status. Figure 14 shows for both income and expenditure poverty, the percentage of those in poverty and not in poverty that are living in households with low work intensity. The figure shows a strong relationship between those with low work intensity and income poverty in all three countries. Additionally, in Germany and Belgium there is also a strong relationship with expenditure poverty, though low work intensity rates are higher for those who are income poor. However, for the UK, there is no evidence of a relationship between expenditure poverty and low work intensity with a negligible difference between the percentages of expenditure-poor and non expenditurepoor with this status. 26

29 4.5 Subjective measures of poverty Figure 15: Difficulty in making ends meet by poverty status, (%) UK GERMANY BELGIUM Income-poor Non income-poor Expenditure-poor Non expenditure-poor Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. Reading note: The figure shows the percentage of those in income and expenditure poverty and those not in income or expenditure poverty that have indicated that they have difficulty or great difficulty making ends meet; for example, for the UK the dark blue bar shows that approximately 35% of the income poor population in the UK state that they have difficulty or great difficulty in making ends meet. EU-SILC also provides a subjective measure of poverty, which can be contrasted with the income and expenditure based measures. Difficulty in making ends meet is defined as someone responding that they lived in a household experiencing difficulty or great difficulty in making ends meet. As with the previous measures examined, the evidence suggests a stronger relationship with income poverty than expenditure poverty, particularly in the UK where the difference in the proportion saying they have difficulty making ends meet between the expenditure poor and non-expenditure poor is relatively small (Figure 15). 27

30 5. Characteristics of those in expenditure poverty Table 5: Comparison of the characteristics of the expenditure poor with the non expenditure poor (% population) UK Germany Belgium Characteristic Expenditure poor Non expenditure poor Expenditure poor Non expenditure poor Expenditure poor Non expenditure poor Gender Male Female Age group Under = Marital status Never married Married Widowed Divorced Household type 1 adult adults > 2 adults adult with dependent children adults with dependent children > 2 adults with dependent children

31 UK Germany Belgium Characteristic Expenditure poor Non expenditure poor Expenditure poor Non expenditure poor Expenditure poor Non expenditure poor Household size 1 person people people people people More than 5 people Activity status Employed Unemployed Retired Other inactive Highest educational level achieved ISCED levels ISCED level ISCED levels ISCED levels Source: EU-SILC : EU-SILC Users database; HBS : Eurostat/ONS. The table compares the characteristics of the expenditure-poor with the non expenditure-poor for the three countries examined. In general, the characteristics of the expenditure poor are more distinctive in Belgium and Germany than in the UK, where there seems to be relatively little difference in the characteristics of those who are expenditure poor and those who are not. In Germany and Belgium, perhaps unsurprisingly, expenditure poor households are characterised by a higher proportion of unemployed householders, though this pattern is not seen in the UK. Across the board, a lower percentage of the expenditure poor are headed by householders in employment than non-expenditure poor households, though the difference is least striking for the UK. Single adult households with children also make up a larger proportion of the expenditure poor than the non-expenditure poor in Belgium and Germany. In those two countries, people who are expenditure poor are also more likely to live in a household where the head has either never been married or is divorced. 29