• 5.1 Are there differences in cost for different subgroups?
• 5.2 Is Annual Household Income a significant factor?
• 5.3 Propensity to incur personal taxi costs taking affordability into consideration.

Page 32

Only 109 of the survey respondents indicated they incur a personal cost for taxis to mitigate their costs of blindness, but only 99 of these gave a dollar value. We estimate that this represents an estimated total of 6206 members, or approximately 55% of the RNZFB membership. The first series of analyses looked at which groups of people within the RNZFB membership incurred this cost. We found that:

1. Based on our estimates, the subgroup of vision impaired, over 65, female, European, not employed and living in a metropolitan area are estimated to represent 14.9% of the total membership (Exhibit A4), but make up 18% of all members with this cost (Exhibit C2.2). Similarly, the subgroup of vision impaired, over 65, female, European, not employed and living in a provincial location are estimated to represent 7.6% of the total membership (Exhibit A4), but make up 13.8% of all members with this cost (Exhibit C2.2).

2. Approximately 39% of all RNZFB members with this cost are thought to be vision impaired, over 65, female, European and not employed (Exhibit C2.2). This subgroup (with and without cost) represents over 30% of the estimated membership (Exhibit A2) and we have estimated that approximately 70% of this subgroup have a personal cost for short distance taxi travel. This is noticeably different to the 55% figure given for the total RNZFB membership.

3. Just over half (3119/6206) of RNZFB members with this cost have a known annual household income of $10,000 to $29,999. While just over 41% (2553/6206) of RNZFB members with this cost live alone. An estimated 29% (1793/6206) of members with this cost live alone and have a known annual household income of $10,000 to $29,999.

Page 33

Exhibit F2.1 provides a boxplot and descriptive statistics of the available personal weekly short distance taxi travel costs. This shows that 75% of these personal costs are less than $20 a week and 75% of costs are greater than $2.31 a week. Without considering any subgrouping, the mean personal short distance taxi cost is $14.52 for the survey respondents.

Exhibit F2.1 Box plot of survey respondents personal cost (left) and Descriptive Statistics (right) of personal weekly short distance taxi travel cost

Descriptive Statistics:

Minimum 0.23

Lower Quartile 2.31

Median 6.92

Upper Quartile 20.00

Maximum 100.00

Mean 14.54

Total of 98 respondents

Note: These statistics are not weighted to the total RNZFB membership population.

Page 34

Exhibit F2.2: Descriptive Statistics weighted to the RNZFB membership

Minimum 0.23

Lower Quartile 2.31

Median 6.92

Upper Quartile 20.00

Maximum 100.00

Mean 14.49

Note: Costs weighted using Integer survey weights

5.1 Are there differences in cost for different subgroups?

A variety of different subgrouping options were considered that used a selection of the six main demographic variables. The following six options have been explored for understanding the differences in amount, and propensity to incur this particular personal cost:

Option 1: Using each individual demographic variable separately

Option 2: Using Visual Acuity and Age

Option 3: Using Visual Acuity and Employment Status

Option 4: Using Employment Status, Gender and Visual Acuity

Option 5: Using Location, Age, Employment Status and Gender

Option 6: Using Location, Age, Employment Status and Visual Acuity

5.1.1 Option 1 using individual demographic variables separately

When individual demographic factors are considered independently of one another, we estimate that:

Page 35

1. The propensity of having a personal short distance taxi travel cost for females (0.62) is significantly higher than males (0.43). This suggests that females are much more likely to incur a personal cost for short distance taxi travel than males. (shown in Exhibit D1.1)

2. There are no significant differences between the mean short distance personal taxi travel costs for males and females. The lower quartile, median and upper quartile for both genders are similar to each other and also to those when no subgrouping is used. (see Exhibit F3.1)

3. The propensity of having a personal short distance taxi travel cost for members between 18 and 65 years old (0.60) is significantly higher than those members not within this age range (0.53). (see Exhibit D1.2)

4. There are no significant differences between the means for the two age groups' short distance personal taxi travel costs. The lower quartile, median and upper quartile for both age groups are similar to each other and also to those when no subgrouping is used. (see Exhibit F3.2)

5. The propensity of having a personal short distance taxi travel cost for members who live in metropolitan or provincial locations is the same (0.64). However, the propensity of having a personal cost for members who live in rural locations (0.31) is significantly lower than those members in the other two locations. It must be noted the lack of taxi services in rural areas is likely to exaggerate the significance in this analysis. (see Exhibit D1.3)

6. The lower quartile, median and upper quartile for all locations are similar to each other and also to that when no subgrouping is used. However, the upper quartile and mean for the rural location is smaller than the other two locations. Exhibit F3.3.1 provides a boxplot for the personal costs for each of the locations. This shows that both non rural locations have several large outliers (costs that

Page 36

are much larger than the rest of the group costs and shown by a * in the boxplot). The differences in location group means for this cost were found to be insignificant. (see Exhibit F3.3.1 for more detail)

7. The propensity of having a personal short distance taxi travel cost for members who are non NZ European (0.58) is significantly higher than those members who are NZ European (0.54). (see Exhibit D1.4)

8. The lower quartile and median for both ethnicity groupings are similar to each other and also to that when no subgrouping is used. However, the upper quartile is larger for the non NZ European group. Exhibit F3.4.2 provides a boxplot for the personal costs for each ethnicity group. This shows that the NZ European group has several large outliers. While the non NZ European group has no outliers, but has a much greater spread of the largest 50% of incurred costs. There are no significant differences between the two ethnicity grouping means for short distance personal taxi travel costs. (see Exhibit F3.41 for more detail)

9. The propensity of having a personal short distance taxi travel cost for members in employment (0.69) is significantly higher than those members not in employment (0.53). Note that this cost is for personal use and does not include the transport costs for getting to and from the workplace. (see Exhibit D1.5)

10. There are no significant differences between the means for short distance personal taxi travel costs for the groups based on employment status. The lower quartile, median and upper quartile for both groupings are similar to each other and also to those when no subgrouping is used. (see Exhibit F3.5 for more detail)

Page 37

11. The propensity of having a personal short distance taxi travel cost for members who are vision impaired (0.56) is significantly higher than those members who are blind (0.53). (see Exhibit D1.6)

12. There is no significant difference between the mean short distance personal taxi travel costs based on visual acuity. The lower quartile, median and upper quartile for both age groupings are similar to each other and also to that when no subgrouping is used. (see Exhibit F3.6 for more detail)

Exhibit F3.3.2 Box plot of personal costs by location

Page 38

Exhibit F3.4.2 Box plot of personal costs by ethnicity subgrouping

5.1.2 Option 2 using Visual Acuity and Age

The propensity of having a personal short distance taxi travel cost of the three subgroups Blind & between 18-65 (0.55), Blind & not between (0.52) and Vision Impaired & not between (0.54) are all similar. However, the propensity for the subgroup Vision Impaired & between 18-65 years old (0.62) is significantly higher than the other three subgroups. (see Exhibit D2.1)

If the subgroups with similar propensity are combined into one subgroup (Blind or Vision Impaired and not between 18-65 years old), the estimated propensity for a cost of members in this group is 0.53.

5.1.3 Option 3 using Visual Acuity and Employment Status

The propensity of having a personal short distance taxi travel cost of all three subgroups are significantly different from each other, with the subgroup Employed (0.69) having the highest propensity (Vision impaired

Page 39

and not employed (0.54), Blind and not employed (0.69). (see Exhibit D3.1)

There are no significant differences between these three subgroup means for short distance personal taxi travel costs.

5.1.4 Option 4 using Employment Status, Gender and Visual Acuity

When comparing members that are employed with those that are not employed (broken up by gender and visual acuity into four groups): The propensity of having a personal short distance taxi travel cost is significantly different for each of the five subgroups (see Table 3 below). The subgroup of members who are employed (0.69) has the largest propensity for a cost. (see Exhibit D4.1 for more detail)

Table 3 Propensities to incur a personal cost for taxi use with subgroups based on employment status, gender, and visual acuity

Exhibit F6.2 provides a boxplot for the personal costs for each of the subgroups.

The personal costs for the male, blind and not employed subgroup appears to have a larger spread in each of the quartiles, in particular the

Page 40

largest 50% of the data. However, the means of the different subgroups are not significantly different.

5.1.5 Option 5 using Location, Age, Employment Status and Gender

When we split the RNZFB membership up into six groups based on location, employment, age, and gender we find that the propensity of having a personal short distance taxi travel cost for the Metropolitan, employed subgroup (0.78) is significantly larger than any of the other subgroups. (see Exhibit D5.1 for more detail) The only pair of groups in this analysis that do not have significantly different propensities to incur personal taxi costs are the two subgroups Provincial & between 18-65 (0.54) and Metropolitan, male & not employed (0.55).

Table 4 Propensities to incur a personal cost for taxi use with subgroups based on location, age, employment status and gender

There are no significant differences between these six subgroup means for short distance personal taxi travel costs.

Page 41

5.1.6 Option 6 using Location, Age, Employment Status and Visual Acuity

The propensity of having a personal short distance taxi travel cost is significantly different for all of these subgroups with the exception of the subgroups Metropolitan, not employed for both blind (0.61) and vision impaired (0.62). Suggesting that for members who live in metropolitan locations and are not in employment that visual acuity makes no difference for their propensity to incur a personal cost. (see Exhibit D6.1 for more detail)

Table 5 Propensities to incur a personal cost for taxi use with subgroups based on location, age, employment status, and visual acuity

If the visual acuity is removed from the subgroups then the estimated propensity for members living in metropolitan locations and not employed is 0.62. (see Exhibit D6.2 for more detail)

Page 42

Table 6 Propensities to incur a personal cost for taxi use with subgroups based on location age, and employment status

There are no significant differences between these six subgroup means for short distance personal taxi travel costs.

5.2 Is Annual Household Income a significant factor?

While it is not possible to go into great detail about the inability to show that those households with higher incomes face less pressure to choose between costs of blindness, it is possible to show the impact income has on the amount spent and propensity to incur personal taxi costs. Note that the four income groups greater than $60,000 have very small numbers of respondents in the sample and have been combined into one annual income group of '$60,000 upwards'.

Page 43

Exhibit D7.2 Estimated Propensity of having a cost by Annual Household Income groups

Exhibit D7.3 Bar chart of Annual Household Income groups estimated Propensity of having a cost

As seen in the tabulated results (Exhibit D7.2) and graph (Exhibit D7.3) the propensity to have a personal short distance taxi travel cost appears to increase with annual house income until the level of income is $30,000 or more. This result needs to be related to our finding that there are

Page 44

predominantly more RNZFB members with annual household income less than $30,000.

No significant differences in the mean amount spent on personal use of taxis could be found between the groups based on household income level.

5.3 Propensity to incur personal taxi costs taking affordability into consideration.

This analysis uses the information that comes from those responses for not using taxis more than the current level as "too expensive", for those people who do not currently use taxis or incur a zero personal cost. There is an indication that these people do have a true personal cost for taxi, but no incurred cost. There are now a total of 132 respondents who do have or would have costs for personal taxi use if they could afford to use them. On these grounds, we can now estimate that there are at least 7208 RNZFB members who want to use taxis for personal use to mitigate the effects of blindness. This equates to 64% of the RNZFB membership. Note that this is an increase from 55% or 6206 members.

The estimated propensity to incur a personal cost has also increased for every individual subgroup explored in Section 5.1, which illustrates that constrained personal taxi costs is widespread across the RNZFB membership rather than being of concern for just particular demographic subgroups.

Table 7 Adjusted propensities for subgroups based on Gender

Page 45

Females still have a statistically significant higher propensity to incur cost, even though the difference has reduced.

Table 8 Adjusted propensities for subgroups based on Age

The difference in propensities for the two age groups has increased, and the between 18-65 group still has significantly higher propensity to incur cost than those outside this age range.

Table 9 Adjusted propensities for subgroups based on Location

Propensities for both the Provincial and Metropolitan subgroups have increased by the same amount, and there is therefore still no difference between these two subgroups. The Rural group still have significantly lower propensity to incur a cost than either the Provincial or Metropolitan subgroups and this difference has increased.

Page 46

Table 10 Adjusted propensities for subgroups based on Ethnicity

Non Pakeha / NZ European still have significantly higher propensity to incur cost, due to the fact that the difference in propensities has remained much the same.

Table 11 Adjusted propensities for subgroups based on Employment Status

The difference in propensities has remained the same between those in employment and those not in employment, and employed members still have significantly higher propensity to incur cost.

Table 12 Adjusted propensities for subgroups based on Visual Acuity

Vision Impaired members still have a significantly higher propensity to incur cost, as the difference in propensities has remained much the same.

Page 47

Table 13 Adjusted propensities for subgroups based on Visual Acuity and Age

There are now significant differences in propensity between the first three subgroups. While the last two subgroups both between 18-65 no longer have distinct propensities but are significantly larger than the other two subgroups. Note that this is a change from the analysis presented in Section 5.1 and that this analysis is preferred.

Table 14 Adjusted propensities for subgroups based on Visual Acuity and Employment Status

All three subgroups still have significantly different propensities after considering affordability, with the group of members who are Employed being the most likely to incur a personal cost for taxis.

Page 48

Table 15 Adjusted propensities for subgroups based on Employment Status, Gender and Visual Acuity

After considering affordability the two 'Not employed, male' subgroups are no longer distinct, but they are still less likely to incur a personal cost for taxis than the other subgroups. The other three subgroups still have distinct propensities, with Employed being the most likely subgroup to incur a cost.

Table 16 Adjusted propensities for subgroups based on Location, Employment Status and Gender

After considering affordability the four subgroups (both Provincial and Metropolitan, and both employed and not employed) the propensities are no distinct. The propensity is now larger for the group 'Provincial, between 18-65' than the group 'Metropolitan, not employed, male' , however this difference is still not significant. The group 'Metropolitan, not employed, male' is still less likely to

Page 49

incur a cost than either of the 'Metropolitan, not employed, female' or 'Provincial, not between 18-65' subgroups.

As a consequence of adding the effect of those people who cannot afford taxis into the analysis, we found that Option 6 (Using Location, Employment Status and Visual Acuity) is now similar to Option 5. It is advisable to use the simpler scenario in future.

Table 17 Adjusted propensities for subgroups based on Annual Household Income

The propensities after considering affordability have all increased. The estimated propensity for RNZFB members with known annual household income of less than $10,000 to incur a cost has increased to over 0.5. The propensity to incur a cost is now distinct for the two known income groups over $30,000, with those members we believe to have an income between $30,000 to $59,999 having the largest propensity.

Page 50

Table 17 Adjusted propensities for subgroups based on Employment status of working aged RNZFB members

Members of working age who are employed still have a significantly higher propensity to incur cost, even though the difference in propensities has reduced.

It is clear that there is an effect on the estimated propensities to incur a personal cost for taxis when data for those individuals with a zero incurred cost who said taxis were "too expensive" are taken into consideration. The adjusted propensities are preferred over those in Section 5.1, because they more accurately reflect the proportions of RNZFB members that have a desire to mitigate their costs of blindness through expenditure on taxis.

Return to the Contents Page