Notice the difference: the impact of intervention on insurance renewals

Finding the cost of your home or motor insurance increase at each renewal is something we are all familiar with. It is also a practice that has come under scrutiny from both the Competition and Markets Authority and the Financial Conduct Authority (FCA) in the UK.

The FCA is currently considering new rules that would constrain these price increases. But for the last couple of years, the FCA’s strategy has been to try and increase the transparency of renewal price increases. The purpose of this has been to encourage customers to switch or negotiate better prices. An extensive evaluation of this intervention has now been published, and the results have been mixed.

In this article, we explore whether these mixed results reflect the limits of what demand-side remedies can achieve, or if the intervention may have benefited from stronger enforcement, with a clearer assessment of the constraints of competition relative to the desired market outcomes. The positive effects the FCA has found, especially reductions in renewal prices, suggest that there is untapped potential to build on this intervention and improve customer outcomes.

The UK’s Financial Conduct Authority (FCA) published its evaluation of this intervention in October 2019. The intervention was introduced in 2017 to encourage general insurance customers to switch. It required providers to show customers last year’s premium alongside the renewal premium. This follows concerns that those customers who do not regularly switch or renegotiate their insurance products face paying significantly higher prices over time compared with customers who regularly shop around.

The Figure below illustrates the broad dynamics of pricing and customer retention within many insurance products. Prices tend to rise over multiple renewals, often flattening out after five or so renewals. Many customers switch regularly, but many more remain for years and a sizable minority can be on the same policy for a decade or more.

Figure 1: an illustration of how insurance pricing tends to rise with customer tenure

Source: Frontier

Differences in prices paid by customers can add up. The FCA has previously found that prices for customers who have been with the same provider for five years can be 70% more than the equivalent for a new customer. They also find that around 1.5 million customers have been with the same insurer for ten or more years, typically paying some of the highest prices for their insurance.

Nudging customers

The intervention the FCA deployed is straight out of the behavioural economics toolkit. It requires providers to display last year’s premium in renewal letters for home, motor and pet insurance. The hope was that showing the increase in prices would encourage customers to either switch or negotiate a better price for their insurance.

The intervention came into force in April 2017 and the evaluation looked at its impact over the following year and a half. The findings are mixed and surprising. The FCA found:

  • a small but positive impact on customers switching and renegotiating in pet and motor insurance but a negative impact in home insurance; and
  • a reduction in the difference in renewal prices in home and motor insurance and an increase in pet insurance.

Overall, the FCA found that customers may have saved between £39m and £330m, driven to a significant extent by the reduction in renewal prices in home and motor insurance.

There are a few notable things about these findings:

  • Markets can dynamically respond to interventions, in this case positively. The FCA has argued that the lower renewal price is due to providers taking action in response to the intervention (presumably lowering renewal prices to reduce the likelihood that customers would switch).
  • The effect of an intervention can be hard to anticipate. The effect on switching was expected to be greater, not least because the FCA had conducted a randomised control trial in 2015 which showed a significant increase in switching and negotiation. In fact, the 2015 trial showed the largest positive effect in switching in home insurance and no effect in motor insurance, almost the opposite of what the real world intervention has achieved.
  • Implementation matters. Providers were given relative freedom to implement the intervention and there was a large variation in how well they did it. Prescriptive interventions are not always the answer, but variation in how well the intervention was implemented might explain some of the disparity in impact between the trial and the intervention.

Upping the ante?

Taking a step back from this particular intervention, the FCA remains concerned about insurance pricing and is now in the process of a market study. The interim report to the market study, published alongside the evaluation of the renewal transparency intervention, sets out a wide range of additional options. Some represent potentially radical interventions, such as stopping automatic renewal of insurance policies, effectively forcing customers to shop around for a new product, or introducing direct controls or restrictions on insurance pricing models.

The current presumption appears to be that the demand-side intervention has not worked well enough, so something stronger is needed. That may be true, but it is not obvious that this is definitely the case. Any demand-side intervention looking to change customer behaviour needs to be designed and implemented within the constraints of competition economics and with an eye on what market outcomes are being targeted. In that spirit, we see a few potential issues with the design and implementation of the intervention:

  • Implementation was poor and more consistent implementation might have led, and could still lead, to stronger effects. The evaluation included scoring the renewal notifications sent by several providers between 0 to 100 based on how transparent the information was. However, the transparency of most providers did not improve and in some cases actually declined. Many providers also scored below 50 out of 100. The table below shows the results of this assessment. More powerful effects might arise from better design, implementation and enforcement of the intervention.

Figure 2  Transparency improvements were limited and scores were low

  • The framing of renewal prices might have underplayed savings to customers. Premium increases tend to accumulate over several years. So while the intervention might have shown customers their home insurance premium has risen £20-£30 in the last year, actual savings from shopping around could be much higher. Switching takes time and effort and this framing may have unintentionally discouraged some customers from shopping around if they felt it was not worth their while.
  • The intervention may not have worked at all for customers who have the most to save. Customers paying the highest insurance premiums are often those who have been with their provider for many years. Often what happens is they see premium increases for 5-6 years and then a stable, but high, premium. Under the intervention the sizable number of customers who have remained with the same provider for 10 or more years may have seen only small differences between last year’s price and this year’s, despite having the most to save from shopping around.

In finding a practical way to implement the intervention, the opportunity to change the behaviour of those customers with the most to gain may have been lost. Alternative designs can be envisaged that target this group. For example, requiring providers to provide a quote for what a new customer would pay. There are challenges to such an implementation, but it is not clear that the potential to do it has been ruled out.

The positive dynamic effects identified by the FCA hint that refining the design of the intervention could drive even better outcomes. At the same time, it is important to ensure a robust exploration of the dynamic effects within the economic framework of the market. To that end, the evaluation leaves a few important questions unanswered. These include:

  • Who benefited from the intervention? Were reductions universal, or did providers target reductions based on, for example, the size of the renewal price increase? If reductions were targeted, it suggests that longstanding customers with limited price rises at renewal, but paying the highest premiums, might have received little or no benefit. Indeed, relatively speaking they may be made worse off as prices fell for other customers with shorter tenures.
  • Did prices increase for new customers? It is possible that providers may have increased prices for new customers to balance the fact that future renewal prices will rise less quickly than they could before the intervention.
  • Have providers sought other ways to raise revenue, such as through “add-ons”? The sale of the primary insurance product is often not the only source of revenue. Providers will often sell ‘add-on’ insurance (e.g. for personal accidents or breakdown cover) and charge credit interest for monthly billing arrangements. One possible response from providers could have been to attempt to increase the price of add-ons and credit to recover revenue.

These questions do not necessarily take away from the fact that the dynamic effects have been a net positive, but they reinforce the need to be clear about what outcomes are being targeted, and working through how the market might respond.

These questions suggest that dynamic effects may be highly sensitive to the design of the intervention. It is possible to imagine designs, like a new business quote at renewal, that may have had the strongest positive effect on switching and renewal prices for customers who are currently paying the highest prices. They also suggest that great care may be needed to assess whether there are less positive unintended consequences, such as higher prices for new customers or for add-on products.

Conclusions

The FCA’s intervention to require the disclosure of last year’s premium is an interesting example of deploying behavioural economics in a demand-side intervention. It shows that trialling is important, but the real world can still surprise, and behaviour change is hard. At the same time, the intervention illustrates the power of unintended consequences, in this case for the better.

We draw several lessons from the intervention and its evaluation:

  • Implementation and enforcement is important and there is value in the FCA considering a stronger approach toward how it polices remedies.
  • Design of the intervention may have lost sight of competitive economic constraints and the desired market outcomes. It seems like focus should have been on encouraging customers paying the highest prices to switch or negotiate. In reality, by only showing last year’s premium the intervention may have failed to provide motivation for such customers to shop around.
  • Careful consideration also needs to be given to how market dynamics play out. Some of that has already been explored in the evaluation, but there is value in going further. For example, looking at who has benefited and whether prices have risen for some customers (e.g. new business) and whether providers have sought to raise revenue in other ways (e.g. raising the price of add-ons).

Given the positive effects the FCA has found, our reflections on implementation and design suggest there could be untapped potential with this intervention. It would be a shame if we do not see what might happen if the FCA enhanced this intervention given all it has learned to date.

The FCA may nevertheless decide that additional intervention is necessary to address concerns around prices paid by longstanding customers. Even so, our reflections on the need for effective implementation and enforcement, and careful consideration of market dynamics and unintended consequences, will be equally, if not more applicable, to whatever new intervention the FCA might introduce.

Notice the difference - the impact of intervention on insurance renewals