Do KPIs Really Improve Public Service?

A government hospital reports treating 20,000 patients in a month. A public school proudly announces a 95 per cent examination pass rate. A land administration office claims to have disposed of 90 per cent of applications within the prescribed time. These figures suggest impressive performance. But what if patients received inadequate treatment, students passed without acquiring essential skills, or land applications were disposed of without resolving citizens’ problems? Can these institutions still be considered successful? This raises a fundamental question in public administration: Are governments measuring what really matters, or simply what is easy to measure?

Nowadays, Key Performance Indicators (KPIs) have become an important instrument of public sector management across the globe. Governments use measurable targets to assess institutional performance, improve efficiency and strengthen accountability. The underlying assumption is that measuring performance encourages improvement. Yet experience and academic research suggest that the relationship between performance measurement and better public services is considerably more complicated. The rationale behind introduction of performance standard and indicator was understandable. Governments needed to demonstrate that public expenditure produced meaningful results. Rather than reporting how much money was spent, public institutions were increasingly expected to explain the actual result. 

Well-designed KPIs can undoubtedly improve public services. A passport office can monitor processing times to identify delays. Hospitals can examine waiting times, infection rates and patient outcomes. Schools can track attendance, learning progress and dropout rates. Such information enables managers to identify weaknesses, allocate resources and evaluate whether corrective measures are working. However, measuring administrative performance is not necessarily the same as improving public service quality. This distinction becomes particularly important when numerical targets begin to dominate institutional priorities.

The problem is commonly explained through Goodhart’s law: when a measure becomes a target, it may cease to function as a reliable measure. Public organisations under pressure to achieve specified targets may concentrate on improving reported figures rather than addressing underlying problems. Consider a hospital required to treat a certain number of patients daily. To meet the target, medical staff might shorten consultations or prioritise uncomplicated cases. Patient numbers increase, but treatment quality may suffer. Similarly, a school evaluated primarily through examination results may concentrate on students most likely to pass while neglecting those requiring additional support.

These are potential consequences of poorly designed incentives, not inevitable outcomes of performance measurement. Nevertheless, they illustrate how institutions can achieve numerical success without delivering corresponding improvements. Research by Gwyn Bevan and Christopher Hood (2006), examining target-based performance management in England’s National Health Service, identified the risks of gaming and distorted organisational behaviour. Their analysis showed why reported improvements should not automatically be interpreted as evidence of better services.

The lesson is significant as KPIs can reveal whether targets have been achieved, but they cannot independently establish whether public institutions have fulfilled their wider responsibilities. For Bangladesh, this debate is particularly relevant. The government introduced the Annual Performance Agreement (APA) system to strengthen results-oriented administration and institutional accountability. Ministries and departments are expected to demonstrate progress against agreed performance indicators.

This represents an important shift towards measurable administrative responsibilities. However, the effectiveness of such arrangements ultimately depends on whether indicators capture meaningful improvements in public services. Consider a district office reporting successful completion of its annual targets for meetings, inspections, training programmes and application disposal. These achievements may demonstrate administrative activity. They do not necessarily establish whether citizens experience fewer delays, reduced procedural difficulties or more transparent services.

The distinction between outputs and outcomes is therefore essential. Governments can become highly efficient at producing measurable administrative outputs without becoming equally effective at solving public problems. Another weakness of conventional KPI systems is their limited attention to citizens’ experiences. A public office may report that 95 per cent of applications were processed within the official deadline. Yet applicants might still have made repeated visits, incurred substantial expenses or struggled to obtain accurate information. From the institution’s perspective, the target has been achieved. From the citizen’s perspective, the service may remain unsatisfactory.

This difference is especially important for disadvantaged communities. A digital service may reduce average processing times while creating difficulties for citizens with limited internet access or digital literacy. Aggregate performance figures can conceal inequalities in service accessibility. The government performance should be assessed through the value public institutions create for society, including service quality, fairness, legitimacy and public welfare. A social protection programme, for example, should not be judged exclusively by the number of beneficiaries enrolled. Its effectiveness also depends on whether eligible households receive assistance reliably and whether the programme contributes to their welfare.

How, then, can governments make KPIs more meaningful? First, performance frameworks should distinguish activities, outputs and outcomes. Indicators must capture the actual purpose of public services rather than merely record administrative accomplishments. Besides, institutions should combine quantitative indicators with citizen feedback, independent evaluations and service-quality assessments. Numbers are useful, but they rarely tell the complete story. Additionally, performance comparisons should recognise differences in institutional resources and operating conditions. A remote hospital facing severe staffing shortages cannot reasonably be assessed without considering its circumstances. Moreover, performance management should encourage organisational learning rather than simply reward success and punish failure. When officials fear consequences for missing targets, they may become reluctant to report problems honestly. A credible system must allow institutions to identify weaknesses and explain the reasons behind them.

For Bangladesh and other developing countries, the challenge is not to abandon KPIs but to improve what they measure and how the information is used. Performance management should shift attention from administrative compliance towards the actual experiences and welfare of citizens. The purpose of KPIs is not to make government statistics look better. It is to make government work better. The most important question for any public institution is therefore not simply whether it has achieved its annual targets, but whether citizens’ lives have improved because of its work.

________________________________

The writer is a public policy analyst