Understanding the Incentives of Commision Motivated Agents in Life Insurance: Theory and Evidence from Indian Life Insurance

Principal Investigators: Shawn Cole (Harvard Business School), Santosh Anagol (Warton School, University of Pennsylvania)
Research Team: Anand Kothari
LEAD Centre: Centre for Microfinance
Focus Area: Insurance
Project Geography: Gujarat, Tamil Nadu
Partner:
Status: Completed

Background:

Financial products in developing countries are primarily sold through agents who are compensated through sales commissions. In an environment where customers may lack the information and financial literacy to make informed product choice decisions, agents are expected to serve as advisors as well. In this study, researchers ask whether motivating agents with commissions is the best way to incentivize them to sell products that are welfare-maximizing to the consumer. The market for life insurance in India is the ideal testing ground for this hypothesis for several reasons;1) The market is very large, with approximately 44 billion dollars in premiums collected in 2008, 2) all sales are executed through the commissions motivated agent channel and 3) it is a complex product, and consumers require help in making decisions.

Two types of life insurance products exist in India, term life insurance and whole life insurance. Term insurance is a pure insurance product which covers risk for a specific term. Whole life insurance products cover risk but also have a committed savings device built into them. These products cover the client’s risk of death for life and provide a cash payout at age 80 based on the amount that has accumulated throughout the premium paying years. The premiums for whole products are substantially higher than those for term products, and for the client, term insurance is almost always a better purchase decision. The amount of money saved in paying premiums for a term product rather than a whole product, if invested in a fixed deposit account yields a substantially higher savings balance over the period of comparison. In general, a term insurance product is better than a whole product if clients have the discipline to save. However, whole insurance in the Indian market is much more popular than term insurance.

In this study, researchers explore why people prefer whole insurance over term insurance when the latter may yield large benefits. One possible reason for the popularity of whole insurance may be the incentives scheme insurance companies provide their agents. In other words, it is possible that insurance companies and sales agents oversell whole insurance since agents earn higher commissions on sale of whole products. However, there could be a few other reasons why consumers make this choice; such as loss aversion and clients’ inability to measure compound interest. The study tests possible explanations as to why people make sub-optimal purchase decisions.

Methodology and Research Design

Researches designed a field and lab experiment to test behavioural factors that lead to purchase decisions. The field experiment, a randomized controlled trial, takes the form of an audit study where CMF researchers trained a few individuals to be mystery shoppers and meet with life insurance sales agents to get recommendations on life insurance products. In this part of the study, customers / mystery shoppers are randomly assigned a script to use in their meeting with the agents. These scripts explain the profile of the customer CMF employees are role playing.

This audit was conducted in both Ahmedabad as well as Chennai in 2010 and 2011 respectively. In Ahmedabad, five such mystery shoppers completed 229 audits (meetings with life insurance agents to collect data without the agent’s knowledge). Their profiles were randomised from a set of 5 scripts which varied the mystery shopper’s level of sophistication, preference for term insurance, whether or not they had met the competition, and whether or not they should ask about ULIP disclosure. In Chennai, the 4 mystery shoppers hired by CMF completed 594 audits where their profiles varied based on 8 scripts which differed on: whether customers were biased towards one of the products, whether customers were actually suited towards one of the products, and whether customers had visited other agents before. The objective was to understand whether agents recommended products that were suitable to the customer or not.

In addition to this experiment, researchers will conduct a lab experiment that will test consumers’ behavioural biases. The experiment, also a randomized control trial, invites people to attend a session where they watch a randomly assigned video and respond to a short survey about their preferences. Participants will be randomized between two videos: the first is a control video, which contains information about types of life insurance products and its characteristics. The control video gives participants reasons why they should purchase life insurance. The other (treatment) video provides more truthful information about the economic advantages of term insurance over whole insurance.

Initial findings

Researchers have found that agents do provide misleading advice to clients, recommending products that are not suitable for the customer. In most cases, the products agents recommend earn them a higher commission. Even in cases where the customer expressed a preference for term insurance, most agents advised against it and instead recommended whole life insurance products by giving customers misinformation about term insurance.

Study Status

The audit phase of the intervention is complete and researchers are working on piloting the lab experiment. For this phase of the experiment, researchers hope to collect data from approximately 600 participants. The average participant will be a middle-aged, middle-income working professional. Data collected from this group will help researchers understand the extent to which information can influence purchase decisions.

Related Resources
WORKING PAPER – Understanding the Incentives of Commissions Motivated Agents: Theory and Evidence from Indian Life Insurance