“Costs avoided” – how we calculate savings


As we mentioned in our last post, one difficulty faced in early years prevention is that money needs to be spent on interventions right now, but the pay-offs often don’t materialise until well into the future. This has implications for research: to show whether an intervention is good value for money, we have to follow people for a long time. This is difficult for many reasons: it places a lot of demands on research participants; there are high costs associated with carrying out such a long study and it takes a long time until results are available.


A potential solution: “costs avoided”

Here’s how we get around these problems in ‘Preventonomics’. Instead of conducting long-term studies, we pull together the existing evidence on the costs associated with childhood problems in the three “A Better Start” areas (nutrition, social and emotional learning, speech and language development) in an economic model. This gives us an estimate of the cost of not intervening.

We then look at interventions shown to be effective in preventing these problems, and calculate how much could potentially be saved if they were implemented. This means we calculate the savings from intervening early as “costs avoided”.


Parenting programmes – an example of a modelling study

One example of this approach is our study looking at savings from parenting interventions to prevent persistent conduct disorders. The full article can be found here. We compared two scenarios: in the first scenario, a “typical” five-year old with conduct problems did not receive the intervention we were looking at. We reviewed the literature and put together the best available estimates on the chance that this five-year old would still have problems by age 16 – this was 50%.

We then calculated the costs over and above those incurred by a child without conduct problems. Next, we estimated the chance that the intervention would further reduce the chance that conduct problems persist. In our analysis, this was an additional 34%. Other factors we considered were drop-out and the chance that the intervention only worked for a short time and problems re-surfaced.

Overall, we showed that an investment of under £1,000 to pay for parenting programmes lead to savings of more than £16,000 over 25 years. Group-based parenting programmes therefore seem like a good investment.


“Guesstimates” vs the “real world”

Of course there are downsides to this approach. The estimates are just that – a best guess, based on data already available to us. “Real world” results will vary, based on factors such as fidelity of implementation, staff qualifications and the motivation of people participating in interventions. On the other hand, these estimates are a good indication of whether it is generally worth taking a chance on a particular type of intervention.


If you would like to read more examples using the “cost avoided” approach, we recommend you take a look at the PSSRU publication “Mental Health Promotion and Prevention: The Economic Case” (free pdf download).


Next time, we will discuss the types of “costs” we will take a look at in the ‘Preventonomics’ work, so please check back here on April 16. In the mean time, follow us on Twitter (@preventonomics) or send us an e-mail ()!

To follow news of the Big Lottery Fund’s ‘A Better Start’ initiative, follow @BIGFirstYears and #ABetterStart.