New measure of catastrophic health expenditures with application on rural Egypt

ABSTRACT

Reducing Out-of-pocket (OOP) health payments is an essential element to protect households from financial risks and eradicate extreme poverty. This paper aimed to provide a new approach for measuring catastrophic health expenditure (CHE) by redefining subsistence spending to comprise the minimum acceptable level of all necessities. The new approach defined OOP health expenditures as catastrophic payments if they undermine a household’s ability to maintain essential needs that are not limited to food needs as assumed by the Capacity to pay (CTP) approach. The equivalence scale is estimated to reflect the dominant consumption patterns and the Cost of Basic Need Approach (CBN) is employed to measure the subsistence spending. We estimated the burden of OOP health payment on household living standards in rural Egypt using data from Household Income, Expenditure, and Consumption Survey in 2015. Rural areas exhibited a considerable incidence of catastrophic payments. Redefining subsistence spending better detected the incidence of catastrophic health expenditure among poor households. The distribution of CHE became highly regressive and demonstrated that catastrophic payments are more concentrated among poor households. Although poor households have incurred relatively small health payments compared to wealthy ones, this has threatened their standard of living. The highest incidence rates of CHE were also observed among uneducated, unemployed, female, elderly heads, among households with chronically ill members, and those who experienced outpatient services. The poor performance of the health insurance system in protecting vulnerable groups from catastrophic payments creates a critical need to redesign its policies and improve its service quality.

KEYWORDS:

• Catastrophic health payments
• capacity to pay
• health insurance
• subsistence spending
• nonfood needs

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 In the growing literature, due to the lack of standard threshold, many studies have used varying thresholds to investigate the robustness of results, they compute results for a wide range of thresholds that remained uniform across different income levels and substantiated that The estimates are sensitive to the threshold used (Akazili et al., 2017; Arsenijevic et al., 2013; Buigut et al., 2015; Chuma & Maina, 2012; Dorjdagva et al., 2016; Experiments, 2008; Ghosh, 2010; Kimani et al., 2016; McHenga et al., 2017; Narcı et al., 2015; Sirag & Mohamed Nor, 2021).

2 Xu (2005) explained that the reported food spending is less than the actual one due to the non-inclusion of the subsidies and self-production, though food expenditures in most surveys include all own produced products and gifts/in-kind received food.

3 To further investigate the validity of Pal’s measure of catastrophic payments (2012), we estimated the expenditure elasticities of different nonfood groups using double log models. Double log models have the property that the coefficients are the elasticity values. We found interesting evidence that the estimated expenditure elasticities of all nonfood commodities groups are more than 1, which means, based on pal's methodology, all nonfood needs are not necessities in rural Egypt. This is contrary to the truth; many households in poor areas, especially rural Egypt, are still unable to meet their nonfood necessities; hence the increase in income has driven Engel’s curves for nonfood groups to increase constantly (All results of double log models are available upon request). This reinforces our discussion that using elasticity estimates to determine basic nonfood needs will give rise to strong disagreements on the appropriate list of nonfood necessities.

4 Engel stated that there are needs whose satisfaction is considered the base for physical sustenance as nourishment, clothing, housing, heating, and lighting and health which stand at the top. While intellectual and spiritual care, legal protection, and public safety, public provisions come in second-order needs, defined as higher order needs (Chai & Moneta, 2012).

5 For sensitivity analysis, Buigut et al. (2015) estimated household-scale multiplier$\sigma$ based on consumption data in Kenya to adjust subsistence expenditure. The study demonstrated that using estimated value by Xu $(\sigma =0.56)$ underestimated the capacity to pay and therefore overestimated slightly the incidence of catastrophic payments in Kenya (Buigut et al., 2015).

6 There is another version of scaling-up procedure, it based on measuring average total expenditure of households whose food expenditure are equal to food poverty line. But we prefer the first version in measuring the basic needs of households as it permits identifying the most important nonfood needs of households living below the poverty line.

7 Standard approaches of incidence and intensity indices of catastrophic health payments and changes in poverty rates due to health expenditures are described in details in Wagstaff and Doorslaer (2003).

8 The calculation of payment gap index was restricted to households with capacity to pay $ct{p}_h^{\ast }$ more than zero

9 All results are weighted to obtain national figures.

10 Nonfood expenditure is the total expenditures on nonfood and nondurables and excludes total expenses on durables; also excludes actual and imputed rentals for housing.