A Question of Choice

AUTHOR: Deborah Tarrant   DATE: 06.09.07   ISSUE 1, 2007

Can consumers really trust the advice of experts – tradespeople or professionals – when choosing which products or services to purchase? How much do experts allow their own interests to sway the advice they give? Do they really have consumers’ concerns at heart?

Professor Denzil Fiebig, an econometrician in the School of Economics, has heartening news for all. According to his studies in consumer choice to date, more often than not we can rely on the recommendations of experts.

This research is relevant to marketers for product design and for determining the emphasis of advertising campaigns.

Illustration: Ron Monnier

Professor Fiebig uses stated preference experiments and economic modelling to compare the trade-offs people use when making purchasing decisions, with what influences the experts when making a recommendation.

His studies provide answers to marketers’ dilemmas. For example, when purchasing a water heater, is the customer more influenced by the cost, the capacity or the warrantee?
Why do plumbers tend to recommend gas water heaters over electrical ones?

How would women and GPs react to changes in the recommended screening interval for pap smears?

His research in the field of health economics has drawn a significant $6,875,000 grant from the National Health and Medical Research Council, along with an Australian Research Council Linkage Grant to study what influences general practitioners and women in contraceptive choice. Professor Fiebig also shares an Australian Research Council Discovery Grant for work on Bayesian choice modelling.

While his research is relevant to marketers, not only for product design but also for determining the emphasis of advertising campaigns, his methodology – which uses sophisticated statistical design theory to model the drivers of consumer choices – provides a template for others wishing to conduct research.

Theoretical economists work on principal agent problems – where people have to use the advice of someone who is expert in the field or rely on someone to make a recommendation. However, he says not a lot of empirical work has been done to test these theories because of the lack of appropriate data. “I have been using stated preference methods to model the interactions.”

While his specialisation is creating individual demand models where people make discrete choices, he is equally interested in survey design or how the information is collected.

“As an econometrician I try to make sense out of data. Quite often real world data won’t behave. No matter how careful you are, sometimes the field or market data you collect won’t allow you tease out the sort of information you want,” he explains.

He also contends with the fact that people send “noisy” signals. “I don’t believe every number I see. I know they are subject to error and I try to minimise the impact of that error in my analysis. When designing surveys, you do the best job you can to minimise those problems.”

In stated preference experiments, respondents are put in a hypothetical yet realistic scenario, and asked to choose a product based on various attributes. Then sophisticated design theory is used to vary the attributes so the real drivers of their choices are uncovered.

For instance, in the case of a water heater is a consumer choosing a smaller capacity water heater simply because it is cheaper or is a plumber recommending a larger one because there’s more profit to be made?

Interestingly, when it came to choosing a water heater, plumbers and consumers make very similar trade-offs, with two exceptions: the type of water heater and the capacity.

“Plumbers had a much stronger preference for gas water heaters than electric and they were more likely to recommend a bigger one. Why? Are they acting in the best interests of the consumer because they know better, or is there some sort of profit motive? These are the questions we wanted to answer.”

Professor Fiebig and his team of researchers found that on capacity, indeed it was more likely that the plumbers had consumers’ interests at heart. However, on the choice of heater, it seems they considered their own convenience.

“Plumbers recommend gas water heaters because they can do the installations themselves. The electric ones involve them having to interact with another tradesperson, an electrician. The consumer, on the other hand, doesn’t care how it gets installed.”

A similar methodology was applied in a research project on how choices were made for cervical screening. In stated preference experiments women and GPs were asked to value attributes when choosing between types of pap smear – a standard test, a potentially more accurate, but more costly, liquid test or the HPV (human papilloma virus) test.

“We had expected the doctors may be more conservative about the sorts of tests they recommended due to the risk of legal proceedings if they hadn’t shown due care – and they showed a heightened concern about under-screening. We thought the doctors may encourage women to screen more often than the women themselves would want to be tested.”

Once again, women and GPs made the same sorts of trade-offs, although there were three exceptions where their decision-making did not coincide. Not surprisingly, women put more emphasis than doctors on the cost. (The cost of the tests was not a factor for the doctors as the remuneration goes to pathologists.)

Another point of difference involved false positive results. There’s the potential for these tests to return false results and where the women returned a “false positive” test, the doctors took a more conservative approach.