Table of Contents Table of Contents
Previous Page  332 / 138 Next Page
Show Menu
Previous Page 332 / 138 Next Page
Page Background






Appendix B

A range of equivalence scales is applied based on different

technical assumptions about economies of scale in consumption.

Among these scales, the most commonly used are the square root

scale, the (old) OECD scale, and the OECD-modified scale

(Burniaux et al., 1998; Förster, 1994; Hagenaars et al., 1994;

OECD, 2012). Poverty rates are affected by different equivalence

scales. For sensitivity tests, I replicated the analyses using the

OECD-modified scale; the first household member is assigned a

value of 1, of 0.5 to each additional adult and of 0.3 to each child.

The results are shown in Appendix B. In general, the equivalized

income is higher if the economies of scale in consumption are

higher (i.e. the value for elasticity is smaller). For example, in a

five-member family with fixed household income (say, 60,000 NT

dollars per month), families containing more adults (e.g.

households in East Asia) will have lower equivalized income (and

thus, higher poverty rates) than households with fewer adults if

greater weight is given to each additional adult in the household.

Thus, compared to the youth poverty rates shown in Table 2 (i.e.

based on the square root scale), the results here (i.e. based on the

OECD-modified scale) show slightly smaller differences in youth

poverty between Taiwan and Scandinavian countries where

intergenerational coresidence is much less common.

However, the application of different equivalence scales does

not substantially affect the ranking of poverty across countries.

This is also observed in the replicated analyses revealed in Table

Appendix 1. Additional analyses are conducted to examine the

correlation of poverty disparities and the three structural effects

based on the two scales. The Spearman’s rho values are very high

(all pair-wise values are around 0.97,


<.001, n=22).

Consistent with the analyses using the square root scale, the

countries with higher levels of youth poverty are Denmark,

Norway, the US, Germany, and Spain while those with lower rates

include Australia, Czech Republic, Japan, Luxembourg, South