Time use value enlarged equivalence scales François Gardes, Université Paris 1 Panthéon-Sorbonne, Paris School of Economics, Crest Imen Sayadi Université Paris 1 Panthéon-Sorbonne, Centre d’Economie de la Sorbonne Christophe Starzec Université Paris 1 Panthéon-Sorbonne, CNRS, Centre d’Economie de la Sorbonne October 23, 2007 Abstract

Introduction In the classic equivalence scale-child cost models only market consumption costs are taken into account. The parental time spent on child care and education in substitution of its other alternative use (market work, leisure, domestic production) is completely ignored. Even if the time use value as an element of child cost has been recognized in the recent theoretical literature very few empirical work has been done up today. These parents’ time inputs can be very substantial and should be taken into account in the same way as the monetary expenditures. Apps (2000)) suggests, that limiting the analysis to the monetary aspects only neglects the key issue of the child cost. Observing the time allocated to the different form of child care observed in the time use surveys especially for young children, it can be reasonably supposed, that the time « expenditure » value could be much larger than specific child market consumption. Thus it seems necessary to define a “full” child cost including both market goods consumption and the monetary value of the parental time spent on the child care and education. We define the “enlarged equivalence scale” and “the full child cost” as the expression of the total, monetary and non monetary expenditure depending respectively on family structure or on the presence of a child. In this paper we propose an original method to estimate the enlarged equivalence scales based on the statistically matched files of the French INSEE Family Budget (2000) and Family Time Use surveys (1998). The estimation of monetary and non monetary costs may help to identify the underlying well-being (which must be supposed to be constant between two different households which are compared in the estimation, a crucial identification problem in the estimation of equivalent scales discussed by Pollack and Wales (1979) and Blundell and Lewbel (1991)1. Second, the usual methods do not take into account the price substitution effects caused by the presence of children (as discussed by Barten, 1964). According to the Barten’s model, children modify the relative prices of goods: one litre of milk is more expensive for the family of three than for the family of one or two children. The complete costs depends changes between households, as these cost are computed using an estimated opportunity cost which differs between households. A difference between the complete prices may also come from different households productivities (measured by the time taken by households to perform some activity). Thus, it may be possible to estimate price effects in the paired expenditures and time surveys. Third, these statistics may contain less zero expenditures, and the enlarged income variable may be closer to a permanent concept (if 1

Blundell-Lewbel showed that it was possible to obtain any cost of child from a cross-section data in the frame of a theoretical model compatible both with arbitrary hypothesis and observed behaviour.

1

some substitution exists during the life-cycle between the two components of income). Fourth, the information is available at an individual (intra-household) level in the time survey, so that a collective model of household behaviour could be tested more easily than with monetary information aggregated by household. The article is organized as follows. Section 1 presents Engel’s and Prais-Houthakker’s specifications of equivalence scales and the econometric methods used to estimate the opportunity cost of time. Section 2 contains the results for the French INSEE surveys.

Section 1. The cost of children and equivalence scale estimations 1.1 Classic monetary scale estimation specifications (Engel, Prais-Houthakker) Engel’s specification Following the Engel law, the changes in food budget share is used as an indicator of household’s standard of living. The food share should increase when a child arrives and this change is considered as a decrease in the well being. The estimation of the cost of children consists in regressing food share on income (or total expenditure) and family size and structure . As a consumption function we will use the Almost Ideal Demand System proposed by Deaton and Muellbauer (1980). For the household i, the functional form of the budget share on food, wf, is: wf = α + β ln x +γ* f(n)+ζ . v + u

(1)

with wf the budget share for food, x the total household income, f(n) = n or ln(n) (or another function of the number of children) and ζ various socio-economic variables. The equivalence scale (ES) for a couple with one child with respect to a childless couple is obtained by a formula: ES =exp (-γ*/β) The underlined hypothesis is that food is a necessity and that the budget share for food increases with the number of children. The method should produce a positive estimate of the cost of children, based on a positive estimate for γ* and a negative one for β. Prais-Houthakker’s specification Its main advantage is to allow the calculation of specific equivalence scales for all categories of expenditures composing the total budget. The total expenditure scale (or income scale) is obtained as a weighted average of specific scales with budget coefficients as weights. The scale coefficients are calculated in this model by estimating the following demand functions describing the changes in the family consumption as a function of its demographic structure at a fixed level of well being: ⎛ x ⎞ qi = g i ⎜⎜ ⎟⎟ mi ⎝ m0 ⎠

2

Where m0 is the total expenditure (income) scale and mi the specific one for the item i. For the the reference household they are set to one. The demand functions are derived from the Working-Leser (Almost ideal) specification of the Engel Curve. The enlarged equivalent expenditure ei on the item i writes: ei / mi = [ Ai (x/m0)log (x/m0)+ Bi (x/m0)]

(2)

Dividing by x and multiplying by mi leads to the formulation in terms of budget shares. Adding socio-demographic variables z and normally distributed error term vectors we obtain the following estimation equation: wi = mi /m0[ Ai log (x/m0)+ Bi ] +∑δs zs +εi . For all items i and for the total expenditure we have the specific and the total expenditures scales: mi = 1 + αi * Nchildren + βi * (Nadult –1), m0= 1 + α0 * Nchildren + β0 * (Nadult –1) where nchildren is the number of person aged less then 18 and nadult is the number of persons aged more then 18. We adopted the identifying hypothesis supposing the zero child expenditure on tobacco and alcohol which implies that the alcohol and tobacco expenditures do not change between couples with and without children at the same level of well being.

1.2 Enlarged equivalence scales estimation method: Matching individual Time Budget data with Consumption data. The proposed method consists in combining at the individual level the monetary and time expenditures into a common, unique good and services consumption structure. This is obtained by adding identically grouped money and money valued time expenditures from two surveys: Family Budget Survey (INSEE BdF 2001) and Family Time Budget (INSEE BdT 1999). The statistical matching between both surveys was done by a grouping method with age, education, localisation as key variables. This method allows a good treatment of measurement errors and zero expenditure problems. The sample of both surveys was limited to the couples with and without children which gives more than 6500 households from BDF and more than 5100 from BdT. We obtain as a matching result about 40 groups (cells) for which we compute common, comparable mean characteristics. We define 8 types of activities or time use types compatible with the available data both from BDF and BDT. Eating and cooking time (BdT) - food consumption BdF), Cleaning and home maintenance (BdT)- dwelling expenditures (including imputed rent) (BdF), Clothing maintenance (BdT) – Clothing expenditures (BdF),

3

Education time (BdT) – Education expenditures (BdF), Health care time (BdT – Health expenditures (BdF), Leisure time (BdT) – leisure expenditures (BdF), Transport time (BdT)- Transport expenditures (BdF), Miscellaneous (BdT) - Miscellaneous (BdF). For each activity and for each household we compute individually the corresponding time use and then the cell weighted average2. This way we obtain for the comparable cells of both surveys the base information for the full time and money expenditure nomenclature. The addition of both will be possible once the individual time value estimated. 1.3 The domestic activities time value evaluations Different methods can be used to give the value to the time spent on domestic activities. W suppose that this value is equal to the opportunity cost of market work. In this approach each not working individual is given his expected hourly wage rate on the labor market for one hour of non market domestic activity. Working individuals are given their actual hourly rate in the market job. We assume that the time use is perfectly exchangeable between market and non market activities. We test the importance of these assumptions by computing the non market work value applying the uniform hourly wage rate equal to the legal minimum. Moreover we suppose that the remuneration does not depend on the nature of domestic activity nor on the day or period when it is done. We treat the household as a unity where the distribution of monetary and non market budgets is a joint decision of the couple. The potential market wage for not working was estimated separately for man and woman using the two step Heckman method. In the participation equation we used variables describing education, age household type, number of children aged less than 5 . In the wage3, equation we used education, age, and individual’s socio-economic category. For each cell we computed the average of the individually estimated expected wages adjusting 4 for income tax and obtained cell hourly wage rate dividing monthly rate by numbers days and hours 5 6 per day . The mean value of non market activities for each cell was computed as a weighted

average of man’s and woman’s volumes of non market activities7. The cell total value of non monetary activities (opportunity cost CO) can be written as : CO =Wh* %DOMh + Wf*%DOMf.8 With Wh: man’s hourly wage rate in the cell. 2

Weighting was necessary to take into account the survey interview day (week-end or a work day) the wieghts are the proportions of interwiewed in the week (0.74) or during the week-end (0.26). 3 Estimated for individuals working full time only. 4 The income tax is computed for every cell using cell averages (taxable income and number of family shares) le. 5 Supposing 22 working days in the month and taking the average of 7h40, 6.h0 hours per day respectively for man and woman. 6 Working man’s non market activities time volume is about 2h36 per day for 3h53 on average in the case of working women. Thus the participation rate in domestic activities is about 40% for man and 60% for women. These rates are practically identical for all cells 7 Working man’s non market activities time volume is about 2h36 per day for 3h53 on average in the case of working women. Thus the participation rate in domestic activities is about 40% for man and 60% for women. These rates are practically identical for all cells 8 the lower limit for 1 hour is equal to minimum wage hourly rate in year 2000.

4

Wf : woman’ hourly wage rate in the cell. %DOMh : cell’s average man’s participation rate in domestic work %DOMh : cell’s average woman’s participation rate in domestic work Using this formula we can compute the average cost of the time for every cell and all activities as the average time multiplied by the cell hourly opportunity cost9. This way we obtain the cell average of the total budget including both monetary expenditures (BDF) and monetary equivalents of non market activities (BDT)10. An alternative method of time use value evaluation was used in order to test the robustness of the described method. The opportunity cost was set uniformly at the level of the hourly minimum wage rate (SMIC) in 2000. With the obtained the full budgets for all defined expenditures and the total household income including monetary and non monetary components. The enlarged equivalence scales can be estimated using the traditional methods applied to the monetary scales. Section 2. Time use value enlarged equivalence scales 2.1 Enlarged equivalence scales and the full child cost estimation methods: the review of different approaches. The child care and education time use evaluation is conceptually at least as complicated as in the case of the monetary expenditures. Indeed, the child arrival in the household disturbs considerably parents’ time budget: market work time, household work time, leisure and even physiological time. The measurement of this global time redistribution cannot be limited to the simple accounting for activities specifically linked with child presence (caring, playing, reading…) which usually underestimate the time devoted to children. This approach would ignore for example the overcharge in the general domestic works or the time shared simultaneously between children and parents. This is why the more sophisticated methods are necessary like those used in monetary approaches for child cost evaluations. Time cost of children estimation methods are very rare as compared with the very rich litterature on monetary costs evaluations. Among the most interesting, C.Barnet-Versat and O.Ekert-Jaffé (2003) build an objective well being indicator and the time budget comparison between families with different number of children. These two criteria allows, under certain hypothesis, to evaluate to what extent the child presence influences the parents’ time budget. The authors define as a well being indicator parents’ personal time including physiological and leisure time. The reduction in this parental time is interpreted as the utility loss due to the child care and education time or an increase in the domestic works induced by child presence. The parental time is regressed in a linear expenditure system with control variables such as : the period and the week day of the survey interview, parents’ work participation, their social category, diploma, parents’ age and age difference, the share of woman’s income in household’s revenue, woman’s potential income, couple marital status, localisation and citizenship. The variable of interest is the children number and age. The results show that a couple spend in average 50 minutes per day of its own time on their unique child aged 6-14. 90 minutes increase of this time is observed for younger children, and it is close to zero for teenagers aged more than 14. Bittman and Goodin (1998) estimate a time equivalence scale like the classic monetary scales. It is an equivalent of an income scale in the sense that it adjusts the total work time 9

Computed on the yearly basis (the time use survey (EDT) reports only daily volumes). Obtained by matching the identically defined groups (cells) in BDT and BDF.

10

5

taking into account the impact of the family demographic structure on market and non market work time. Time equivalence scale determines the extra non monetary parental work time induced by the child presence. The time child cost can be deduced from this scale. Brudbury (2004) proposes to estimate the full child cost using the adapted Rothbarth’s method which differentiates the goods exclusively consumed by adults and children. The adult’s good consumption is used as a well being indicator. The Bradbury’s approach defines the parents’ personal and leisure time as an adult good. The relationship between this good, the family size and income is used to determine the total child cost. It is defined as the difference between the income levels required to have the same consumption of adults’ good (leisure and personal time) in the household with and without children. He obtains a total child cost which is much higher than the monetary and explains by the fact that in his approach a child can have higher impact on the parents’ standard of living than an extra adult. These studies do not consider the whole cost associated to the family structure, and the substitutions which may exist between monetary and non monetary costs. This is the object of our computation of enlarged expenditures using matched surveys. 2.2 Estimation results and comparison of monetary and enlarged scales The estimation results of the Prais-Houthakker enlarged scales are given in Table 1 for the grouped household units (38 cells) using two time opportunity cost evaluation. The first based on individual wage equation estimation and the second on the minimum wage (SMIC) imputation. Table 1– Enlarged equivalence scales coefficient estimates (Prais-Houthakker method) Time opportunity cost evaluation Budget

Food

Housing

Clothing

Health

Transport

Education

leisure

child 0.39

0.02

0.45

0.32

0.58

1.24

0.26

(0.11)

(0.008)

(0.12)

(0.04)

(0.13)

(0.34)

(0.06)

Adult 0.87

0.19

0.70

0.54

1.74

0.81

(0.23)

(0.006)

(0.28)

(0.07)

(0.30)

alcohol and tobacco 11

Other

Income scale

0.20

0.38

(0.03)

(0.03)

0.60

0.89 0.59

0.68

(0.31)

(0.27)

(0.11)

(0.15)

(0.10)

alcohol and tobacco 1

Other

Income scale

0

SMIC evaluation Budget

Housing

Clothing

Health

Transport

Education

leisure

child 0.32

0.03

0.48

0.37

0.21

0.18

0.33

(0.006)

(0.09)

(0.04)

(0.52)

(0.06)

(0.02)

(0.02)

1.2

0.20

0.61

0.50

0.47 (0.08) 0.99

1.70

(0.13)

1.04

0.71

0.73 0.36

0.64

(0.60)

(0.005)

(0.23)

(0.06)

(0.32)

(0.54)

(0.30)

(0.14)

(0.08)

Adult

Food

0

(0.13)

Estimation on 38 cells defined by education, location and age classes. Data: Family Budget Survey, INSEE BdF, 2001

The first conclusion is that the time opportunity cost evaluation method does not change the results considerably. The computed scales are a little higher for opportunity cost evaluation based on wage equation estimates than for those based on minimum wage imputation.

11

Identification hypothesis supposing zero expenditure for children of this item

6

However the respective adult child coefficient ratios are very close except for transport and education for which the estimates are less precise. The specific scales coefficients for adults are higher than those for childs for all expenditure types except for education where the child coefficient is higher for the two evaluation methods. The table 2 presents the classic monetary expenditure scales estimated on Family Budget Survey (INSEE BdF 2001). Table 2- Classic monetary equivalence scale estimation (Prais-Houthakker method) Budget

child Adult

Food

Housing

Clothing

Health

Transport

leisure

alcool and tabacoo

0

Other12

Income scale

0.25 0.10

0.43

0.34

0.16

0.19

(0.010) (0.005)

(0.03)

(0.05)

(0.02)

(0.01)

0.16

0.18

(0.008)

(0.009)

0.79 0.20

0.55

0.37

0.41

0.43

0.38

(0.03)

(0.06)

0.54 (0.03)

0.30

(0.018) (0.007)

(0.017)

(0.04)

(0.013)

(0.013)

Estimation on 9529 households after income instrumentation. Data: Family Budget Survey, INSEE BdF, 2001

Third, the results show clearly that the enlarged scale coefficients are much higher than those obtained by the classic monetary approach. This is particularly significant in the case of food, transport, leisure and income scales for both methods of time value evaluations. We note, particularly that the child weight doubles for the enlarged income scale when compared with monetary one. Also, the ratio child/adult coefficient computed from the estimated monetary income scales is smaller than that based on estimated enlarged income scales. This implies that taking into account the time use will have more impact on the relative child weight when compared with an adult. The estimation of the Engel model (Table 3) shows also, that enlarged child weights integrating both monetary costs and time use costs (using the two evaluation methods), are higher than those estimated using monetary costs only. It confirms the result obtained with Prais-Houthakker model. However it can be noted that enlarged scales based on Engel model are more sensitive to time value evaluation method. For the two models, child weight is higher for the opportunity cost evaluation method based on wage estimation. Table 3- Enlarged equivalence scales Engel method

Couple without children Couple with one child

Time opportunity cost evaluation 1 1.43 (0.16) SMIC evaluation

Couple without children

12

1

education expenditure is included in other expenditures

7

Prais-Houthakker method 1 1.23 (0.022)* 1

1.27 (0.08)

Couple with one child

1.20 (0.015)*

Table.4- Monetary equivalence scales (Engel Method) Engel method Couple without children

1 1.18 (0.016)

Couple with one child

Prais-Houthakker method 1 1.13 (0.007)*

*bootstrap simulations

Summing up, the obtained enlarged scales high coefficient estimates imply much higher child cost evaluation, than those obtained with traditional monetary approach. The difference in scale coefficients obtained by two methods is larger for children than for adults and is not sensitive to evaluation method as shown comparing Prais-Houthakker and Engel model estimates. Indeed, the child is an important factor of parental time consumption increase, larger in average, than in the case of an additional adult in the household.

3.7 Conclusion The concept of enlarged equivalence scales integrating both monetary costs and the time use costs has been applied by matching family budget (BdF 2001) and time use data (BdT 1998) surveys. The estimated enlarged equivalence scales are much higher than the traditional, monetary ones especially as far as the child coefficient is concerned. The child weight doubles for the income scale, and this result is invariant to the time value evaluation method. The difference between adult’s coefficients between monetary and enlarged income scale is smaller than for child but also not sensitive to the time value evaluation method. Thus, adding the value of different domestic activities to the monetary expenditures results in a strong increase in the estimated child cost and in a smaller increase for an extra adult cost. This result is particularly important for those households which are less constrained by their monetary expenditures or their time use and which can substitute between these two components of the enlarged expenditure. This can be the case for example of wealthy households able to buy services in order to decrease the time component of their non market activities. It may have the strong implications for the inside family monetary and time resources sharing rules – a question which should be developed in the future research.

8

References Barnet-Verzat C. et Ekert-Jaffé O., 2003, Le Coût du Temps Consacré aux Enfants », Communication aux 20èmes Journées de Microéconomie Appliquée. Barten. A.P., 1964, Family composition, prices and expenditure patterns, in Economic Analysis for National Economic Planning, P. Hart, G. Mill et J. Whittaker, Eds., 16th Meeting of the Colston Society, London, Butterworth. Becker G.S., 1965, A Theory of the Allocation of Time, The Economic Journal, vol 75, p.493-517. Bittman M. et Goodin R.E., 1998, An Equivalence Scale for Time, Social Policy Research Centre, discussion paper n°85. Blundel, R. et Lewbel, A., 1991, The Information Content of Equivalence Scales, Journal of Econometrics. Brudbury B., 2004, The price, Cost, Consumption and Value of Children , Social Policy Research Centre, discussion paper n°132. Gardes, F., Starzec, C., 2000, Economies of Scale and Food Consumption: a Reappraisal of Deaton-Paxson Paradox , Income and Wealth Congress. Gardes F, Duncan G J, Gaubert Patrice, Starzec C, 2005, Panel and Pseudo-Panel Estimation of Cross-Sectional and Time Series Elasticities of Food Consumption: The Case of U.S. and Polish Data, Journal of Business & Economic Statistics , vol. 23, no. 2, 242-253. Muellbauer J., 1980, The Estimation of the Prais-Houthakker Model of Equivalence Scales , Econometrica, vol.48, n°1, 153-173. Prais S.J. and Houthakker H.S., 1955, The analysis of Family Budgets, with an Application to two British Surveys Conducted in 1937-39 and their Detailed Results, Cambridge University Press. Tsakloglou, 1991, Estimation and Comparison of Two Simple Models of Equivalence Scale for the Cost of Children, The Economic Journal, vol. 101, n° 405, 343-357.

9

Appendix I The Data French Family Budget Survey (INSEE, BdF 2001) The survey covers all civilian non institutional households in metropolitan France and overseas departments. Data collection is made in several waves (8) during 1 year over two calendar years starting in the spring. The interviewer will visit the household three times. It is necessary to respect the relevant waiting periods between visits in order to let the household fill the diaries. The interviewer is completing the expenditure information by a special “quality questionnaire describing the condition of the data collection. The diaries are kept by households during 14 days. The data collection unit is the household. The household’s expenditures are not recorded over a year. It is thus necessary to have a uniform break down of the sample over the year, in order to take into accounts the seasonal effects that may affect some expenditures. Data collection is made in several waves (8) during 1 year over two calendar years starting in the spring. The interviewer will visit the household three times. It is necessary to respect the relevant waiting periods between visits in order to let the household fill the diaries. The interviewer is completing the expenditure information by a special “quality questionnaire describing the condition of the data collection. The diaries are kept by households during 14 days. The BdF survey being long, difficult, time-consuming, it is not easy to avoid refusals. Even if it is mandatory, the non response rate is rather high. For the three last surveys (1985, 1990, 1995), for instance, out of 100 households, 13 refuse to answer from the beginning, and 12 begin the survey but do not complete it. In total 75% give responses to all parts of the survey.The annual expenditure is estimated by using the questionnaires and the diaries. For several types of expenditures the diary is the only source (eg food expenditures). This is why the quality and reliability of diaries is so important. For other types of expenditures, the questionnaires and the diary are potential sources. Whenever this is the case (for insurances or home renovation) the questionnaire is privileged because it is assumed to be more reliable for regular and big expenditures. The French INSEE Time Budget survey (INSEE, BdT, 1999) A European harmonization made this survey differ from its predecessors: activities have been observed each 10 minutes. Each member of the family note her activity during a day fixed by the survey team and the characteristics of all members aged more than 15 are recorded. The household is visited two times in its home so that the survey does not include households taking their holidays away from home. 8186 households were surveyed and 15441 individuals time records were exploited. 139 activities were considered, with an aggregation, first into 43, and finally into four : biological activities, physiological, domestic labour, leisure time. For multiple activities, a principal and a secondary activity are defined by the individual (if one activity causes another, it is automatically the principal). The two main determinant of time use are the sex of the individual and her presence on the labour market. The average time and the activity rates were weighted to take into account non responses. The surveys distinguishe between average time for the whole population and the average time for the individual who practiced the activity during the recorded day.

10

Appendix II Computation of the time opportunity costs for males and females Male Female Pre-tax Pre-tax wage wage

Total wage

Average Male Female Male Female Male Female Tot Male Fem.% Opp.Co Tax by cell Post-tax Post-tax post-tax post-tax Domesti Domesti Dom % dom dom st /cell wage wage wages/h wages/h c time c time time

C1

93600

49600

143200

1463 92643,7 49093,3

7,0

4,4

2,8

4,5

7,3

0,4

0,6

5,4

C2

98700

54300

153000

1891 97480,1 53628,9

7,4

4,8

2,53

3,38

5,91

0,4

0,6

5,9

C3

110950

51050

162000

3782

108359 49858,2

8,2

4,4

2,3

3,93

6,23

0,4

0,6

5,8

C4

125200

75200

200400

5562

121725 73112,9

9,2

6,5

2,43

3,51

5,94

0,4

0,6

7,6

C5

121700

59700

181400

6165

117563 57671,1

8,9

5,1

3,48

4,96

8,44

0,4

0,6

6,7

C6

130200

59000

189200

7596

124972 56631,3

9,5

5,0

2,65

3,95

6,6

0,4

0,6

6,8

C7

139300

59400

198700

8580

133284 56835,1

10,1

5,0

2,91

4,63

7,54

0,4

0,6

7,0

C8

124900

61800

186700

6066

120841 59792,1

9,2

5,3

2,5

3,31

5,81

0,4

0,6

7,0

C9

173700

73700

247400

14714

163369 69316,7

12,4

6,2

2,88

4,15

7,03

0,4

0,6

8,7

C101 169100

81200

250300

13903

159707 76689,7

12,1

6,8

2,4

3,41

5,81

0,4

0,6

9,0

C112

65200

44000

109200

573 64857,9 43769,1

4,9

3,9

2,35

3,16

5,51

0,4

0,6

4,3

C113

65000

39500

104500

358 64777,3 39364,7

4,9

3,5

2,65

3,51

6,16

0,4

0,6

4,1

C122 104000

54000

158000

1989

102690 53320,2

7,8

4,7

2,46

4,06

6,52

0,4

0,6

5,9

C123 106000

49000

155000

1740

104810 48449,9

8,0

4,3

2,45

4,11

6,56

0,4

0,6

5,7

C131

99500

44500

144000

2639 97676,5 43684,5

7,4

3,9

3,38

5,21

8,59

0,4

0,6

5,3

C132 102600

45500

148100

2918

100578 44603,5

7,6

4,0

2,86

4,76

7,62

0,4

0,6

5,3

C133 104600

42300

146900

2714

102667 41518,5

7,8

3,7

2,28

4,13

6,41

0,4

0,6

5,2

C141 118900

44100

163000

4616

115532 42851,1

8,8

3,8

4,05

5,71

9,76

0,4

0,6

5,9

C142 118400

46400

164800

4873

114899 45028,0

8,7

4,0

3,43

5,4

8,83

0,4

0,6

5,8

C143 122800

49000

171800

5831

118632 47336,9

9,0

4,2

2,08

3

5,08

0,4

0,6

6,2

C211

83200

33100

116300

625 82752,9 32922,1

6,3

2,9

2,55

3,43

5,98

0,4

0,6

4,4

C212

86000

50000

136000

1646 84959,1 49394,9

6,5

4,4

2,6

3,96

6,56

0,4

0,6

5,2

C213

89000

47200

136200

1783 87834,9 46582,1

6,7

4,1

2,81

3,61

6,42

0,4

0,6

5,2

C221 136000

64000

200000

4052

133244 62703,4

10,1

5,6

2,65

4,66

7,31

0,4

0,6

7,2

C222 133000

60000

193000

3678

130465 58856,6

9,9

5,2

2,65

4,63

7,28

0,4

0,6

6,9

C223 140000

57500

197500

4509

136803 56187,3

10,4

5,0

2,56

4,03

6,59

0,4

0,6

7,1

C311 148600

65200

213800

7023

143718 63058,3

10,9

5,6

2,53

3,78

6,31

0,4

0,6

7,7

C312 149600

70200

219800

9005

143471 67324,0

10,9

6,0

2,26

3,2

5,46

0,4

0,6

8,0

C313 144800

74200

219000

9289

138658 71052,8

10,5

6,3

2,25

3,25

5,5

0,4

0,6

8,0

C314 158000

82300

240300

11851

150207 78241,2

11,4

7,0

2,25

2,48

4,73

0,5

0,5

9,1

C321 195500

87500

283000

12799

186658 83542,7

14,2

7,4

3

4,03

7,03

0,4

0,6

10,3

C322 193000

92700

285700

13551

183845 88303,2

14,0

7,8

2,4

3,96

6,36

0,4

0,6

10,2

C323 193000

88400

281400

13062

184041 84296,7

14,0

7,5

2,5

3,63

6,13

0,4

0,6

10,1

C324 199200

97000

296200

11779

191278 93142,6

14,5

8,3

2,25

3,38

5,63

0,4

0,6

10,8

C332 194700

90700

285400

20390

180789 84220,1

13,7

7,5

3,05

3,76

6,81

0,4

0,6

10,3

C333 193200

90200

283400

19295

180046 84058,8

13,7

7,5

2,6

3,83

6,43

0,4

0,6

10,0

C342 185900

75100

261000

17673

173312 70014,8

13,2

6,2

3,36

4,16

7,52

0,4

0,6

9,3

C343 186600

73800

260400

17819

173831 68749,9

13,2

6,1

3,55

4,66

8,21

0,4

0,6

9,2

11