Utilization and Selection in an Ancillaries Health Insurance Market

• Author: Nathan Kettlewell
• Published date: 01 December 2019
• DOI: http://doi.org/10.1111/jori.12250
• Date: 01 December 2019

©2018 The Journal of Risk and Insurance (2018).

DOI: 10.1111/jori.12250

Utilization and Selection in an Ancillaries Health

Insurance Market

Nathan Kettlewell

Abstract

I study two important aspects of the Australian private ancillaries health

insurance (PAHI) market—moral hazard and self-selection. PAHI covers

out-of-hospital health services including dental, optometry, physiotherapy,

and chiropractic. Using instrumental variables methods, I find evidence that

PAHIincreases utilization, particularly for dental, physiotherapy, chiroprac-

tic, osteopathy, and acupuncture. I also find evidence of selection effects—

both adverse and favorable. Several variables jointly predict a person’s

propensity to insure and to utilize health services. There is little evidence

of self-selection based on the joint probability of different health services,

which has important implications for understanding the commercial success

of PAHI.

Introduction

Understanding who purchases health insurance, and how insurance affects decisions

to utilize health services, is important to inform future health policy. Although there

exists a large volume of research on these topics across different institutional envi-

ronments and for different types of health services, no article to date has examined

these issues in detail in the case of Australian private ancillaries health insurance

(PAHI). PAHI, otherwise known as general treatment or “extras” health insurance,

is a supplementary health insurance product that covers a range of out-of-hospital

services excluded by Medicare, Australia’s universal health insurance system. These

include some preventative and discretionary treatments like massage, acupuncture,

and naturopathy. The most quantitatively important service is dental, followed by

optometry, physiotherapy, and chiropractic. Similar supplementary insurance prod-

ucts are available in other developed countries such as Switzerland, Germany, the

Netherlands, and Canada (Colombo and Tapay, 2004).

The first contribution of this article is to estimate the causal utilization effect on the

following out-of-hospital services: dental, physiotherapy, chiropractic, osteopathy,

Nathan Kettlewell is at the School of Economics, University of Sydney,Sydney,NSW, 2006, Aus-

tralia. Kettlewell can be contacted via e-mail: nathan.kettlewell@sydney.edu.au. I am grateful

to Deborah Cobb-Clark, Denise Doiron, and an anonymous referee for helpful comments that

greatly improved this article. The article has also benefited from feedback from participants at

various seminars and conferences. Any mistakes are my own.

1

989

989

. Vol. 86, No. 4, 989–1017 (2019).

2The Journal of Risk and Insurance

acupuncture, optometry, naturopathy,psychology,and counseling. While recent work

has examined the role of PAHI on utilization of dental services (Hopkins, Kidd, and

Ulker, 2013; Srivastava, Chen, and Harris, 2017), this is the first article to consider

other expense categories. A second contribution is to examine the characteristics of

those who purchase PAHI and in particular, the presence and direction of selection

effects. There has been no focused study on selection effects in this market to date.

Health events in the PAHImarket differ from those in most insurance markets, which

usually occur with great uncertainty. For example, 65 percent of dental visits in Aus-

tralia are for preventative check-ups (Chrisopoulos and Harford, 2013). Physiother-

apy, chiropractic, osteopathy, acupuncture, naturopathy, psychology, and counseling

are also more discretionary than other events typically associated with insurance, as

treatment is often nonessential.

The predictability of expenses and discretion afforded to consumers about whether

to actually utilize services suggest that the traditional risk-sharing model may be un-

able to rationalize the PAHI market. For example, without a risk-reducing role, self-

selection is expected to be more strongly related to expected usage (adverse selection)

than other insurance markets, which could create the conditions for complete market

unraveling (Akerlof, 1970). The discretionary nature of expenses should also create

pressure for this market through moral hazard (consumers either taking actions that

increase their probability of making a claim or reacting to the price incentives pro-

vided by coverage). Despite this, the PAHI market is actually more profitable than

many other insurance markets, most notably the private hospital market (gross profit

margins are 11.7 percent for hospital coverage compared to 24.3 percent for PAHI;

PHIAC (2013). Participation is also high—currently around 56 percent of Australians

have this insurance.

The fact that PAHI policies cover a wide variety of health services could provide an

insight into their ongoing commercial success. Under certain conditions, bundling

multiple services can be used to increase consumer participation in markets where

price discrimination is restricted (Thatcher and Clemons, 2000). If risks for different

health services are independent of each other, then as the number of services covered

grows, consumers become more homogeneous in their value of a given policy. Other

research has focused on limitations in consumer psychology. For example, experimen-

tal research suggests that consumers are often narrow sighted when choosing health

insurance. Schram and Sonnemans (2011) find that people often sample information

on health insurance contracts by focusing on a single common feature. Similarly,

Bhargava, Loewenstein, and Sydnor (2017) find that people frequently ignore price

and focus only on coverage when choosing a policy. Such heuristic choice strategies

could have important consequences in the case of PAHI. Policies are typically priced

so that consumers can only receive more in benefits than they pay in premiums if they

utilize multiple health services. It is therefore important that consumers understand

their joint risk of various health events when deciding whether to purchase insurance.

The Australian PAHI system has many desirable features for the applied researcher.

Insurance status is not tied to employment and therefore reflects personal demand.

2The Journal of Risk and Insurance

990