Monte Carlo or bust?

9 Sep 13
Sophisticated forecasting techniques, such as the Monte Carlo method, can help predict future income and demand for services. But which approach is best?

By Manj Kalar | 09 September 2013

Sophisticated forecasting techniques, such as the Monte Carlo method, can help predict future income and demand for services. But which approach is best?

Monte Carlo or bust

Thinking about forcasts usually conjures up images of a crystal ball, accompanied by the stereotypical view of these – what is the point? Forecasts are only going to be wrong, aren’t they? Not so: done properly, forecasting is an important element of the business-planning toolkit.

To manage the significant decrease in public sector finances over the past few years, organisations have, quite rightly, focused on cuts to back-office functions to protect the front line. However, as financial pressures continue, new and more innovative solutions are going to be required to avoid cuts to frontline services.

To understand how to balance ever-decreasing budgets, a good assessment is required of future income and demand for the services. This is where forecasting will help.

All across the public sector, there has been a relentless focus on reducing costs (matching the decrease in government spending allocations) to manage the supply side – ie, the organisation’s ability to provide the level of public service required. However, with warnings of further cuts affecting an organisation’s ability to continue to provide current service levels, leaders need to forecast not only the supply side, but the likely demand for the services.

Managing demand is unfamiliar territory to many, and this change in perspective goes to the heart of the relationship between the state and the individual – from providing public services to the individual at the point of need to assessing which services can be provided. It also suggests inherent uncertainty: do we know what, when and how much public service is needed?

The difficulty that the finance professional has is precisely the inherent uncertainty in forecasts where he/she strives for accuracy. Done well – by choosing the correct forecasting method for the situation – risks can be managed.

Any forecast (and forecasting model) inevitably contains a degree of uncertainty – because it is a plan for the future, and certain assumptions are made – so it is valuable to quantify and understand how much uncertainty there is within a forecast.

The winning entry for the 2013 Sir Harry Page Award, which recognises excellence in public finance and accountancy, exemplifies how an innovative approach has been used to forecast demand (and therefore the impact on expenditure).

The team at the Department for Work and Pensions were recognised at CIPFA’s annual conference for using the Monte Carlo method (a complex statistical technique, not a trip to the casino). This method goes beyond standard forecasting by taking a range of possible values – instead of a single guess – to create a more realistic picture of what might happen.

The key feature of a Monte Carlo simulation is that it can tell you, based on how you create the ranges of estimates, how likely the resulting outcomes are.

This is only one of many approaches to forecasting that might be appropriate in certain circumstances. So how do we know which method is right?

Those who attended CIPFA’s Local Government Accounting Conference in Birmingham in July will have heard Malcolm Prowle (professor of business performance at Nottingham Business School) and Roger Latham (past CIPFA president and visiting fellow at NBS) talk on this topic. They are also speaking at CIPFA’s central government conferences in Liverpool and London this September.

Now they have, with the support of CIPFA’s in-house expertise, written A guide to forecasting methods in public services. The guide provides an overview of the theory and framework of forecasting, and of the different types of qualitative and quantitative techniques. It also sets out which techniques are most suited to which context.

It is not designed just for those working on forecasts, but for senior finance professionals and other decision-makers who rely on forecasting when signing off major spending decisions or policy initiatives. Effective decision-making across the public sector requires assurance on the bases on which key forecasts are prepared.

To fail to accurately assess ever-increasing demand and likely income because it is ‘too difficult’ is no longer an option. To survive, we need to be able to accurately forecast the demand for services to assess whether this demand can be met.

Manj Kalar is CIPFA’s technical manager for central government and financial management. A guide to forecasting methods in public services will be available on the CIPFA website from September.

This feature was first published in the September edition of Public Finance magazine


CIPFA logo

Did you enjoy this article?