A Fuzzy Approach to Financial Literacy Measurement

Albert Hizgilov and Jacques Silber

Introduction

The topic of financial literacy has become quite popular in recent years. Atkinson and Messy (2011) defined it as ‘a combination of awareness, knowledge, skill, attitude and behavior to make sound financial decisions and ultimately achieve individual financial well-being’. Studies dealing with the determinants of financial literacy have shown that it is positively related to income (see Klapper et al., 2015) and education (see, for instance, Calvet et ah, 2009), that women are less financially literate than men (see, for example, Lusardi and Mitchell, 2008; Bucher-Koenen et ah, 2016), that employees and self-employed are more financially literate than unemployed (see Lusardi &c Mitchell, 2011), that financial literacy is lower in rural than urban areas (see Klapper & Panos, 2011), that even religion (Alessie et ah, 2011), political opinions (Arrondel et ah, 2013) and ethnicity (Nejad & O’Connor, 2016) matter.

As far as the impact of financial literacy is concerned, it appears that financial literacy and financial management skills are correlated (see Jappeli & Padula, 2013), that people who are less financially literate use more costly forms of borrowing (see Lusardi & Scheresberg, 2013), tend to use informal sources of borrowing (Klapper et ah, 2012), plan less well for their retirement (Boisclair et ah, 2017), are less likely to invest in stock (Arrondel et ah, 2012), and even when they participate in financial markets, they get a lower return on their investment (Chu et ah, 2017). Finally, financial literacy is also associated with greater wealth accumulation (see, van Rooij et ah, 2012)

Lusardi and Mitchell (2008, 2011) were among the first to attempt to estimate the extent of financial literacy. They focused their attention on three aspects of financial literacy, the notions of compound interest, inflation rates and risk diversification.

In the present chapter we take a multi-dimensional approach to financial literacy measurement. We borrow tools from the literature on multi-dimensional poverty measurement, focusing on the so-called fuzzy approach. We examine, in particular, the impact of the weighting schemes (the weights given to the different aspects or questions related to financial literacy) and emphasise the ‘fuzzy approach’, which is advocated by Rippin (2013) in the context of multi-dimensional poverty measurement.

The chapter is organised as follows. The second section describes the methodology adopted to deal with different weighting schemes, while the third section describes the ‘fuzzy approach’ of Rippin (2013). The fourth section is devoted to an empirical illustration based on Israeli data. Concluding comments are then given in the last section

A Multi-dimensional Approach to the Measurement of Financial Literacy

Let us assume a survey which includes questions on various aspects of financial literacy. We will suppose that for each question the individual can give either a correct or a wrong answer. Let then д;. be a binary variable equal to 1 if individual i gives a correct answer to question /', or equal to 0 otherwise.

In this section we will focus our attention on the selection of weights for the different questions.

Giving the Same Weight to All the Questions

Looking at All the Questions Together

If, as a whole, there are К questions, the proportion of questions to which a correct answer was given by individual h will be expressed as:

If T refers to the total number of individuals participating in the survey, the proportion of individuals having given a correct answer to question k will be:

If we now consider the whole questionnaire, the proportion a of correct answers will be:

Taking a Separate Look at Each of the P Parts of the Questionnaire

Let us now assume that the questionnaire is divided into P parts covering different aspects of financial literacy.

Let aphk be a binary variable equal to 1 if individual b gave a correct answer to question k which belongs to part p of the questionnaire, to 0 otherwise.

The proportion of individuals who gave a correct answer to question k in part p of the questionnaire will then be:

As a consequence, individuals will have given a correct answer to a proportion dp of the questions included in part p of the questionnaire, with:

and where K1' refers to the total number of questions in part p of the questionnaire.

Note then that a in (9.3) may also be defined as:

Giving a Different Weight to Each Question

Analysing the Whole Questionnaire

Using Equation (9.2), and following Cerioli and Zani (1990), Cheli et al. (1994) and Cheli and Lemmi (1995), the weight of question k will be:

so that the weighted proportion of questions answered correctly by individual h will be:

while the weighted proportion of questions to which a correct answer was given in the whole population will be:

Looking at Part p of the Questionnaire

Using Equation (9.3), the weight of question k in part p of the questionnaire will be:

so that the weighted proportion of questions to which individual b gave a correct answer in part p of the questionnaire will be:

while the weighted proportion of questions to which individuals in the whole population gave a correct answer part p will be:

Aggregating All k Parts

Given lp{b) as it was defined in Equation (9.11), we can now derive an alternative estimate X’(h) of the proportion of questions to which a correct answer was given by individual h in the whole questionnaire where:

where is the weight given to part p. This weight w may be defined, in a way similar to that in which we defined in Equation (9.10) the weight of a given question, so that we would write that:

But it is also possible to assume that this weight w will be equal to the share of the number of questions in part p in the total number of questions in the whole questionnaire. It is also possible to give the same weight to each

138 A. Hizgilov and J. Silber

part p, no matter how many questions it includes, in which case we would write that wp = j^J.

No matter which weight is selected, when we proceed in two stages, the alternative measure of the proportion of questions to which individuals in the population gave on average a correct answer will be expressed as:

Applying Rippin’s (2012) Fuzzy Approach to Multi-dimensional Poverty to the Measurement of Financial Literacy

In this section we will ignore the case where financial literacy is measured in two stages and focus our attention on its measurement when we look at the questionnaire as a whole.

Let us start by defining the ‘financial literacy counting function’ flch of individual b as:

where, as before, ahk = 1 if individual b gave a correct answer to question k, to 0 otherwise, whereas тк is the weight given to question k. The choice between various types of weights has been discussed above. In the specific case where the weights are the same for all questions, we would define тк as:

One possibility is to apply Alkire and Foster’s (2011) approach to multidimensional poverty measurement to the case of financial literacy and say that if this weighted proportion of correct answers is smaller than some threshold, the individual will be considered as financially illiterate, while if it is greater than the threshold, he/she will be considered as financially literate.

Rather than taking such a dichotomous approach, we can generalise this method (see Yalonetzky, 2012; Silber & Yalonetzky, 2013). We can, for example, apply Rippin’s (2012) approach to multi-dimensional poverty measurement to financial literacy and define a ‘fuzzy’ identification function flih that requires taking into account the relationship between the various components of financial literacy (see Silber & Yalonetzky, 2013, p. 12). We then write:

Note that the function flih is non-decreasing in flch. Such a fuzzy identification function flih makes thus a distinction between the multi-dimen- sionally financially literate individuals and assumes different degrees of multi-dimensional financial literacy identification (Rippin, 2017). The function flih is considered to be fuzzy because unless flch = 1 or flch = 0, each individual will be identified as being somewhat multi-dimensionally literate (Silber & Yalonetzky, 2013, p. 13). In other words, individuals are identified as having different degrees of multi-dimensional literacy, depending on the number of questions to which they gave a correct answer and on the type of relationship that is assumed to exist between the various components of financial literacy. The shape of the identification function flih depends on the value of//. If 0 1, it has a convex shape. The choice of a specific form will depend on whether it is assumed that the components of financial literacy are assumed to be substitutes or complements. If they are considered as substitutes, we have the case of a concave function. The increase in financial literacy is marginally decreasing in flch. This implies that the higher the number of questions to which an individual gives a correct answer, the more he/she is identified as being financially literate, but this increase in identification becomes smaller and smaller as flch increases. If the components of financial literacy are perfect substitutes, we have the ‘union case’, so that as soon as an individual gives a correct answer to one question, he/she is considered as financially literate.

If the components of financial literacy are imperfect complements, we have the more general case approximated by a convex identification function. Then the higher the number of questions to which an individual gives a correct answer, the more he/she is identified as being financially literate, and this increase in financial literacy rises with flch. In the specific case where these components are perfect complements, as long as an individual did not give a correct answer to all the questions, his/her financial literacy will be equal to 0. This is the intersection case mentioned previously (Silber & Yalonetzky, 2013).

The assumption of a particular relationship among the components of financial literacy is certainly not a simple task. It is often hard enough to determine the degree of complementarity or substitutability on a pair-wise basis, let alone among combinations of К variables taken 3, 4, up to n at a time. We will therefore assume different degrees of complementarity (// = 1.25, 1.50, 1.75) and substitutability (/? = 0.25, 0.50, 0.75) among the components of financial literacy.

As in the poverty measurement literature, the financial literacy identification function identifies the multi-dimensionally literate, but it also needs to take into account the intensity of the multi-dimensional literacy (Silber & Yalonetzky, 2013). However, with ordinal (binary or dichotomised) variables, the multi-dimensional literacy ‘depth’1 cannot be estimated because one cannot define a gap between the degree of literacy of an individual and some literacy threshold. We will therefore assume that the ‘breadth of literacy’ is measured via the number of questions to which the individual gave a correct answer. The individual multi-dimensional literacy function will then be defined as the product of the identification function defined previously and a function that captures the breadth of multi-dimensional literacy. Let b{flch) be the function measuring the multi-dimensional literacy breadth. This function depends on the number of questions to which the individual did not give a correct answer. The degree of multi-dimensional literacy MDLn will then be expressed as:

As mentioned previously, Rippin (2012) assumed that flih (flch) = flctK Rippin also assumed that b{flch) = flch. Therefore, following Rippin, the degree of multi-dimensional individual literacy MDLh will be expressed as:

Then, again following Rippin, the extent of multi-dimensional financial literacy MDL in the whole population will be written as:

It is important to stress, that the measure proposed by Rippin (2012) takes into account the degree of inequality, between the individuals (households) classified as financially literate, in the number of correct answers. Such a property does not hold for the index developed by Alkire and Foster (2011).

The Empirical Analysis The Database:

The empirical analysis uses a survey conducted by Israel’s Central Bureau of Statistics (CBS) in 2012. This representative survey was part of Israel’s Social Survey, which is conducted every year, and it covered 1,171 individuals who were at least 20 years old. There were 105 questions in the questionnaire and, taking into account the definition of financial literacy given by the OECD; we focused our attention on 58 questions that are shown in Appendix 1. This financial literacy survey comprises questions on financial knowledge but also on the debt, credit card and savings behaviour of the respondents. The survey also includes questions on some socio-economic characteristics of the individuals participating in the survey (e.g. age, gender, education, employment status, household income).

This database had been used previously by Meir et al. (2016) who looked at the link between financial literacy and retirement planning in Israel. Hizgilov and Silber (2019) also used this database and estimated regressions where the dependent variables were the overall degree financial literacy or the extent of financial literacy in three domains distinguished in the survey: financial behaviour, financial knowledge and financial attitude. They examined the case where an equal weight was given to the questions as well as when the questions have different weights. But in their empirical analysis they did not take into account the possibility of making a two-stage analysis, where for each of the three domains equal or different weights are given to the various questions, while in a second stage different aggregation procedures are considered to derive an overall measure of financial literacy. Moreover, when applying Rippin’s approach, they did not make a distinction between the case where the different attributes of financial literacy, reflected in the various questions, are assumed to be substitutes and the case where they are supposed to be complements. In this chapter we also examine various degrees of substitutability and complementarity.

The Impact of the Weighting System on Financial Literacy by Socio- Economic Characteristics

Results When Using a One-Stage Procedure

GIVING AN EQUAL WEIGHT TO EACH QUESTION

Table 9.1 gives information on the financial literacy scores of the respondents by socio-economic characteristics and for each of the three types of questions when an equal weight is given to each question.

Financial Behaviour

This category of questions includes the largest number of questions (50 out of 58 questions). It appears that 35.4% of the individuals may be considered as financially literate, as far as financial behaviour is concerned (a score lower than for the two other categories of questions). For men, the score is 36.53%, for women 34.33% (a statistically significant difference).

Financial Knowledge

The score (39.31%) obtained is higher than that observed for the category of questions devoted to financial behaviour. Financial knowledge is higher among men (46%) than among women (33.02%), among Jews (41.57%) than among non-Jews (28.99%), this being true for both genders. As far as age is concerned, the score for financial knowledge is highest in the age group 55-59 (score of 48.06%) and 34-39 (score of 45.33%). For employment status we observe that those who are employed have the highest financial knowledge score (42.75%), while those who are not in the labour force have a sore of 31.74, with the unemployed having a score of 30.85%. The score for financial knowledge rises with the level of education, being equal to 28.22% among those who have up to 12 years of education to 37.54% among those with up to 15 years of education and to 50.04% among those with 16 or more years of education.

Table 9.1 Financial literacy scores by individual characteristics, for each of the three categories of questions, when an equal weight is given to each question

Characteristic

Behaviour

Knowledge

Attitude

Number of participants

Share of population

Overall sample participants

35.40

39.31

47.77

1,213

100

Gender

Men

36.53

46.00

46.34

588

48.47

Women

34.33

33.02

49.12

625

51.53

1 T-test 1

(-2.9861)

(-7.3277)

(1.3007)

Religion and minorii

:ies

Jewish

36.94

41.57

46.63

955

82.03

Non-Jewish

28.38

28.99

52.98

218

17.97

1 T-test 1

(-9.1837)

(-5.4005)

(2.2883)

Employment status

Employed

37.49

42.75

46.61

840

69.25

Not in the labour force

31.78

31.74

50.17

299

24.65

Unemployed

26.32

30.85

51.35

74

6.10

Age

20-24

34.01

28.24

46.67

165

13.61

25-29

35.56

34.08

50

159

13.12

30-34

37.35

41.91

51.51

132

10.89

35-39

35.85

45.53

42.28

123

10.15

40-44

33.58

39.78

46.37

124

10.23

45-49

34.38

43.46

47.19

89

7.34

50-54

35.63

38.20

44.27

96

7.92

55-59

36.98

48.06

46.28

94

7.76

60-64

36.62

43.67

43.89

90

7.43

65+

34.87

39.45

55.34

140

11.55

Education (years)

1,212

Up to 12

29.88

28.22

46.85

397

32.75

13-15

35.16

37.65

49.43

353

29.13

16+

40.35

50.04

47.40

462

38.12

Note: The numbers of questions for each category of questions are respectively: Behaviour = 50; Knowledge = 6; Attitude = 2.

Financial Attitude

Financial attitude is the category of questions where the highest scores were achieved. However, given that this part includes only two questions, we need to be careful when interpreting the results and comparing them with those observed for the other two parts of the questionnaire. The overall score for financial attitude is 47.77%. Interestingly, women have a higher score (49.12%) than men (46.34%) (this difference is not statistically significant). Non-Jews have a higher score (52.98%) than Jews (46.63%) (statistically significant difference). Among non-Jews, Christian Arabs have the highest score (59.37%) while other non-Jewish individuals have a score of 58.75% and Muslim Arabs a score of 50.93%. As far as age is concerned, the highest scores are achieved in the age group 25-34 and among individuals who are over 65. Unemployed have a higher score (51.53%) than individuals who are not members of the labour force (50.17%) and employed individuals (46.61%). Education has no linear relationship with the score on financial attitude: the score of those who have up to 15 years of education is 49.43%, that of those with up to 12 years of education 46.85% and that of those with more than 16 years of education 47.40%.

THE CASE WHERE THE QUESTIONS HAVE DIFFERENT WEIGHTS

In what follows each question may have a different weight. The weights used are those proposed by Cerioli and Zani (1990). As was explained in above, the intuitive idea is that the less successfully answered questions should receive a higher weight while the questions to which most individuals gave a correct answer should have a lower weight. Table 9.2 shows the results obtained

Table 9.2 Financial literacy scores: the questions have different weights

Characteristic

Financial literacy score

Number of participants

Share of population

-test

Overall sample participants

18.20

1,213

100

Gender

Men

19.60

588

48.47

4.8843

Women

16.89

625

51.53

Religion and minorities

Jewish

19.08

955

82.03

6.9221

Non-Jewish

14.16

218

17.97

Employment status

Employed

19.66

840

69.25

Not in the labour force

15.28

299

24.65

Unemployed

13.53

74

6.10

Age

20-24

16.43

165

13.61

25-29

17.33

159

13.12

30-34

19.23

132

10.89

35-39

19.42

123

10.15

40-44

17.33

124

10.23

45-49

18.10

89

7.34

50-54

18.56

96

7.92

55-59

20.11

94

7.76

60-64

19.03

90

7.43

65+

18.00

140

11.55

Education (years) Up to 12

14.03

  • 1212
  • 397

32.75

13-15

17.83

353

29.13

16+

22.08

462

38.12

when adopting such a weighting scheme and when taking into account all the questions in the questionnaire. It then appears that the scores are lower than those observed when giving an equal eight to each question. The overall level of financial literacy then turns out to be quite low (18.2%) for the population as a whole, males having a higher score (a difference of almost 3%, in absolute terms). We observe a higher score among Jews than non-Jews, employed individuals than those unemployed or not belonging to the labour force. The highest levels of financial literacy are achieved in the age groups 30-39 and 55-64 and the score is higher, the more educated the individual is.

Results When Using a Two-Stage Procedure

In what follows we divide the questions into three categories of questions related respectively to financial behaviour, knowledge and attitude, assuming again that the questions have different weights. Table 9.3 gives the scores for each category of questions and various socio-economic and demographic characteristics of the individuals.

To derive the aggregated score obtained we proceeded in two stages, as mentioned above. First within each category of questions we applied the Cerioli and Zani procedure, using Equations (9.7) and (9.8). Then in a second stage we aggregated the three financial scores, using the weighting procedure described in expression (9.14).

It then appears that the overall aggregated financial literacy score is 28.18 when we aggregate all three parts using the Cerioli and Zani weighing system. Similar results are obtained when we use equal weights for each of the three parts (an aggregated score of 32.51%). However, the aggregated financial literacy score significantly declines (19.38%) when we use a weighting system that is based on the share of questions within each part. Whatever the aggregation method we use, men have a higher score than women. Their score is 30.27% when using the Cerioli and Zani procedure, 20.92% when using the share of questions weighting scheme and 34.55% when using equal weights. The corresponding scores for women are, respectively, 26.22%, 17.92% and 30.59%. Jews have higher scores than non-Jews: 28.94%, 20.27% and 33.01% for the three weighting schemes previously mentioned. The corresponding scores for non-Jews are 24.94%, 15.29% and 30.25%. The score is higher among employed individuals than unemployed. More educated people have higher scores. The highest level of financial literacy is observed among those belonging to the age groups 55-59 and 30-34, except for the aggregation using an equal weighting system where those individuals above the age 65 have a high financial literacy score.

Applying Rippin’s Fuzzy Approach to Multi-dimensional Poverty to Financial Literacy

Table 9.4 gives the financial literacy scores by socio-economic and demographic characteristics for each category of questions separately, as well as

Characteristic

Behaviour

Knowledge

Attitude

Weighting scheme

Weighting scheme

Weighting scheme

Number of participants

Share of population

Cerioli and Zani

Share of questions

Equal weights

Weighting scheme

45.45

28.18

19.38

32.51

1,213

100

Gender

Men

17.41

42.66

43.59

30.27

20.92

34.55

588

48.47

Women

15.40

29.18

47.20

26.22

17.92

30.59

625

51.53

1 T-test 1

(3.84)

(7.76)

(1.66)

(4.73)

(5.22)

(3.79)

Religion and minorities

Jewish

17.20

38.04

43.79

28.94

20.27

33.01

955

82.03

Non-Jewish

12.60

25.10

53.04

24.94

15.29

30.25

218

17.97

1 T-test 1

(6.84)

(5.66)

(3.27)

(3.54)

(6.71)

(2.02)

Employment status

Employed

17.71

39.22

44.65

29.67

20.87

33.86

840

69.25

Not a member of the working force

13.71

28.15

46.84

25.01

16.35

29.57

299

24.65

Unemployed

11.87

26.58

48.94

24.10

14.67

29.13

74

6.10

Age

20-24

15.28

24.42

45.12

24.36

17.26

28.28

165

13.61

25-29

15.79

30.21

47.83

26.85

18.39

31.28

159

13.12

30-34

17.25

38.18

49.00

30.10

20.51

34.81

132

10.89

35-39

17.28

42.56

41.03

29.62

20.72

33.63

123

10.15

40-44

15.43

36.09

44.01

27.50

18.55

31.84

124

10.23

45-49

15.99

39.82

43.75

28.79

19.42

33.19

89

7.34

50-54

16.85

35.47

42.28

27.66

19.65

31.53

96

7.92

55-59

17.81

44.47

45.60

31.41

21.52

35.96

94

7.76

60-64

17.10

39.37

40.34

28.47

20.21

32.27

90

7.43

65

16.02

36.03

51.44

29.40

19.31

34.50

140

11.55

Education (years)

1,212

Up to 12

12.61

25.10

44.92

23.18

15.02

27.55

397

32.75

13-15

16.06

34.17

46.29

27.77

18.98

32.17

353

29.13

16+

19.86

45.98

45.36

32.81

23.44

37.07

462

38.12

Note: The numbers of questions for each category of questions are respectively: Behaviour = 50; Knowledge = 6; Attitude = 2

Table 9.4 Financial literacy scores, for each category of questions separately, as well as when no distinction is made between the various categories of questions, using Rippin’s approach for the case of substitutability

Charact.

Behav.

0=0.25

Knowl.

0=0.25

Attit.

/1=0.25

Tot. Fin. Lit. score 0=0.25

Behav.

0=0.5

Knowl.

0=0.5

Attit.

/1=0.5

Tot. Fin. Lit. score 0=0.5

Behav.

0=0.75

Knowl.

0=0.75

Attit.

/1=0.75

Tot. Fin. Lit. score 0=0.75

Gender

Men

29.08

41.40

42.79

30.31

23.10

37.51

39.68

24.48

18.73

34.61

37.47

19.89

Women

26.87

28.23

45.53

27.13

20.99

24.40

42.40

21.38

16.69

21.59

40.16

17.01

1 T-test 1

(-3.15)

(-7.56)

(1.28)

(-4.37)

(-3.25)

(-7.68)

(1.26)

(-4.53)

(-3.37)

(-7.72)

(1.24)

(-4.57)

Religion and minorities

Jewish

29.36

36.96

43.05

30.10

23.28

33.11

39.91

24.19

18.81

30.27

37.67

19.56

Muslim

20.34

21.91

47.52

20.89

15.18

18.04

44.54

15.79

11.59

15.23

42.41

12.08

Arabs

Christian Arabs

25.81

27.31

55.87

26.37

20

23.44

52.81

20.69

15.75

20.62

50.62

16.44

Other

24.37

30.82

54.95

25.52

18.87

26.45

51.62

20.15

14.95

23.25

49.25

16.07

Non-Jew

Employment status

Employed

29.89

37.89

43.07

30.67

23.78

33.86

39.98

24.71

19.27

30.88

37.77

20.03

Not a member of the working force

24.49

27.41

46.31

25.02

18.85

23.89

42.94

19.52

14.81

21.31

40.53

15.39

Unemployed

19.74

26.61

48.54

20.76

14.77

23.20

46.08

15.80

11.24

20.72973

44.32

12.16

Age

20-24

26.58

23.30

43.66

26.27

20.73

19.46

41.03

20.60

16.45

16.65

39.15

16.24

25-29

27.97

29.13

46.23

28.23

21.91

25.09

42.92

22.34

17.50

22.14

40.57

17.80

30-34

29.76

36.73

47.76

30.64

23.67

32.52

44.47

24.70

19.16

29.45

42.12

20.04

35-39

28.38

41.06

39.02

29.48

22.37

37.30

36.18

23.60

18.02

34.53

34.15

19.03

40-44

26.20

34.76

42.69

27.16

20.39

30.70

.39.48

21.45

16.21

27.73

37.18

17.14

45-49

27.12

38.89

42.88

28.29

21.33

35.01

39.10

22.65

17.09

32.10

36.40

18.29

50-54

28.26

34.23

40.69

28.79

22.42

30.92

37.55

23.09

18.15

28.48

35.31

18.76

55-59

29.67

43.37

43.13

30.98

23.84

39.45

40.37

25.27

19.49

36.52

38.40

20.62

60-64

29.08

39.08

40.42

30.03

23.06

35.03

37.39

24.13

18.62

32.04

35.22

19.52

65+

27.36

35.29

51.53

28.52

21.49

31.79

48.18

22.74

17.18

29.16

45.79

18.34

Education (years)

Up to 12

22.79

23.85

43.43

23.05

17.32

20.38

40.43

17.69

13.45

17.82

38.29

13.70

13-15

27.58

32.70

45.78

28.19

21.56

28.68

42.59

22.35

17.17

25.72

40.31

17.81

16+

32.67

45.28

43.77

33.89

26.41

41.20

40.58

27.78

21.72

38.19

38.31

22.92

Note: The numbers of questions for each category of questions are respectively: Behaviour = 50; Knowledge = 6; Attitude = 2.

the scores when no distinction is made between the categories when using Rippin’s approach in the case of substitutability.

It appears that for all values of /?, the overall financial literacy level of the individuals is below 30%, which is quite low. The highest scores are achieved when /? is equal to 0.25 and, as expected, the scores are decreasing with /?. We also observe that men outscore women when no distinction is made between the categories as well as in each category separately, except for financial behaviour. Also, those gender differences are statistically significant in all cases but financial behaviour. Furthermore, the scores are higher among Jews than among the other religious population subgroups, both when we do not make a distinction between the different categories of questions and separately for each category of questions. The only exception concerns financial behaviour where the population subgroup ‘Other non-Jews’ scores higher than Jews. Moreover, we observe that those individuals who are employed score higher than those who are unemployed or not members of the labour force. Additionally, we observe that the higher the educational level, the higher the scores of the participants, this being true when no distinction is made between the different categories of questions, as well as when we look separately at each of the three categories of questions. Looking at the various age groups, we observe that the highest scores are achieved by those who are between the ages of 30-39 and 55-64.

The results for the case of complementarity are very similar to those concerning the case of substitutability.2 The highest score observed occurs when /? is equal to 1.25. This score is slightly lower than 12% and declines with /?. As previously, we observe that men outscore women and that Jews score higher than other population subgroups. We also see that individuals who are employed are more financially literate than those who are unemployed or those who do not belong to the labour force. Furthermore, the higher the education level of an individual, the higher his/her financial literacy score. As far as age is concerned, the highest scores are observed for the age groups 30-39 and 55-64.

Concluding Comments

In this chapter a multi-dimensional approach to financial literacy measurement was proposed, with special emphasis on the so-called ‘fuzzy approach’. Particular attention was given to the possibility of using various weighting schemes (the weights given to the different aspects or questions related to financial literacy) and on their impact on the estimation of the overall degree of financial literacy. The chapter also stressed the ‘fuzzy approach’, that was advocated by Rippin (2013) in the context of multi-dimensional poverty measurement, in particular the fact that the various attributes of financial literacy that are supposed to be reflected in the different questions may be considered as ‘substitutes’ or ‘complements’.

The empirical illustration was based on data collected in a survey conducted by Israel’s CBS in 2012. This empirical analysis showed that, whatever the weighting scheme that was adopted, the aggregated financial literacy score was higher among males, that the relationship between age and this aggregated score was U shaped, and that married individuals had a higher aggregated financial literacy score. Other things remaining constant, Muslim Arabs generally have a lower aggregated score, this being also true for the unemployed. Finally, ceteris paribus, the higher the educational level, the higher the aggregated financial literacy score.

The same kind of conclusions was drawn when focusing on the Rippin approach, and these conclusions did not depend on whether the attributes of financial literacy were substitutes or complements and the degree of substitutability or complementarity did not really affect the direction of the impact of the various socio-economic variables that had been taken into account.

Notes

  • 1 The ‘depth’ of poverty measures how far the poor are, on average, from the poverty line.
  • 2 The detailed results are available upon request from the authors.

References

Alessie, R., van Rooij, M., 6c Lusardi, A. (2011) ‘Financial literacy and retirement preparation in the Netherlands’, Journal of Pension Economics and Finance 10(4): 527-545.

Alkire, S., 6c Foster, J. (2011) ‘Counting and multi-dimensional poverty measurement’, Journal of Public Economics 95(7-8): 476-487.

Arrondel, L., Debbich, M., 6c Savignac, F. (2012) ‘Stockholding and financial literacy in the French population’, International Journal of Social Sciences and Humanity Studies 4(2): 1309-8063.

Arrondel, L., Debbich, M., 6c Savignac, F. (2013) ‘Literacy financial literacy and financial planning in France financial literacy and financial planning in France’, Numeracy 6(2): Article 8.

Atkinson, A., 6c Messy, F.-A. (2011) ‘Assessing financial literacy in 12 countries: an OECD/INFE international pilot exercise’, Journal of Pension Economics and Finance 10(4): 657-665.

Boisclair, D., Lusardi, A., 6c Michaud, P.-C. (2017) ‘Financial literacy and retirement planning in Canada’, Journal of Pension Economics and Finance 16(3): 277-296. Bucher-Koenen, T., Lusardi, A., Alessie, R., 6c van Rooij, M. (2016) ‘How financially literate are women? An overview and new insights’, Journal of Consumer Affairs. https://doi.org/10-l 111/joca.12121

Calvet, L.E., Campbell, J.Y., 6c Sodini, P. (2009) ‘Measuring the financial sophistication of households’, American Economic Review 99(2): 393-398. Cerioli, A., 6c Zani, S. (1990) ‘A fuzzy approach to the measurement of poverty’, in C. Dagum 6c M. Zenga (Eds.), Income and Wealth Distribution, Inequality and Poverty: Proceedings of the Second International Conference on Income Distribution by Size: Generation, Distribution, Measurement and Applications, Held at the University of Pavia, Italy, 28-30 September, Vol. 19 (pp. 272-284). Berlin: Springer Berlin Heidelberg.

Cheli, B., Ghellini, A., Lemmi, A, Sc Pannuzi, N. (1994) ‘Measuring poverty in the countries in transition via the TFR method: The case of Poland In 1990-199Г, Statistics in Transition 1(5): 585-636.

Cheli, В., Sc Lemmi, A. (1995) ‘“Totally” fuzzy and relative approach to the multidimensional analysis of poverty’, Economics Notes by Monte dei Pascbi di Siena 24(1): 115-134.

Chu, Z., Wang, Z., Xiao, J.J., 5c Zhang, W. (2017) ‘Financial literacy, Portfolio choice and financial well-being’, Social Indicators Research 132(2): 799-820.

Hizgilov, A., Sc Silber, J. (2019) ‘On multi-dimensional approaches to financial literacy measurement’, Social Indicators Research, forthcoming, https://doi.org /10.1007/sl 1205-019-02227-4

Jappelli, T., 5c Padula, M. (2013) ‘Investment in financial literacy and saving decisions’, Journal of Banking & Finance 37(8): 2779-2792.

Klapper, L.F., Lusardi, A., Sc Panos, G.A. (2012) Financial Literacy and the Financial Crisis. NBER Working Paper No. 17930. http://www.nber.org/ papers/wl7930

Klapper, L., Lusardi, A., 5c Van Oudheusden, P. (2015) Financial Literacy around the World: Insights from the Standard & Poor Ratings Services Global Financial Literacy Survey. http://gflec.Org/wp-content/uploads/2015/l 1/Finlit_paper_16_ F2_singles.pdf

Klapper, L.F., 5c Panos, G.A. (2011) ‘Financial literacy and retirement planning: The Russian case’, Journal of Pension Economics and Finance 10(4): 599-618.

Lusardi, A., De, C., Sc Scheresberg, B. (2013) Financial Literacy and High-Cost Borrowing in the United States. NBER Working Paper Series No. 18969. https:/ /doi.org/10.1017/CB09781107415324.004

Lusardi, A., Sc Mitchell, O.S. (2008) ‘Planning and financial literacy: How do women fare?’ American Economic Review 98(2): 413-417.

Lusardi, A., Sc Mitchell, O.S. (2011) ‘Financial literacy and retirement planning in the United States’, Journal of Pension Economics and Finance 10(4): 509-525.

Meir, A., Mugerman, Y., Sc Sade, O. (2016) ‘Financial literacy and retirement planning: Evidence from Israel’, Israel Economic Review 14(1): 75-95.

Nejad, M.G., Sc O’Connor, G. (2016) ‘An intersectional approach to evaluating consumer financial literacy’, Journal of Financial Services Marketing 21(4): 308-324.

Rippin, N. (2012) Distributional Justice and Efficiency: Integrating Inequality Within and between Dimensions in Additive Poverty Indices. Discussion Papers, 128. Courant Research Centre, ‘Poverty, Equity and Growth in Developing and Transition Countries; Statistical Methods and Empirical Analysis’, Georg- August-Universitat Gottingen.

Rippin, N. (2013) Considerations of Efficiency and Distributive Justice in Multidimensional Poverty Measurement. Doctoral Dissertation. Germany: George-August-Universitat Gottingen, https://www.die-gdi.de/uploads/media/ Rippin_Efficiency_and_DistributiveJustice_Online_Publication.pdf.

Silber, J., Sc Yalonetzky, G. (2013) ‘Measuring multi-dimensional deprivation with dichotomized and ordinal variables’, in G. Betti Sc A. Lemmi (Eds.), Poverty and Social Exclusion: New Methods of Analysis, Routledge Advances in Social Economics (pp. 9-37). New York: Rourledge. van Rooij, M.C.J., Lusardi, A., 6c Alessie, R.J.M. (2012) ‘Financial literacy, retirement planning and household wealth’, The Economic Journal 122(560): 449^178.

Yalonetzky, G. (2012) ‘Conditions for the most robust multi-dimensional poverty comparisons using counting measures and ordinal variables’, Social Choice and Welfare 43(4): 773-807.

 
Source
< Prev   CONTENTS   Source   Next >