• Does government need more staff or a better system?

A government officer retires. The chair becomes empty. What happens next? In many institutions, the answer is almost automatic: fill the vacancy. But that simple response may soon become one of the biggest mistakes in public sector workforce planning. 

The real question is no longer, ‘Who should sit in this chair next?’ It is, ‘Will this chair even need a person five years from now?’

Divergence of workload and workforce

For decades, government workforce planning followed a familiar pattern. How many officers are retiring? How many vacancies exist? How many recruits are needed? How many additional posts should be created? 

That logic made sense when public administration depended heavily on people. Files moved because people carried them. Applications were checked manually. Registers were updated by hand. Letters were typed, copied, signed, stamped, and filed. Citizens stood in queues because almost every government transaction required a human being somewhere in the process.

However, that world is changing fast. Artificial Intelligence (AI), automation, digital platforms, electronic records, online payments, and data analytics are steadily taking over tasks that once required large numbers of employees. This creates a difficult but unavoidable question for every government department: are we planning tomorrow’s workforce using yesterday’s work? If so, the result may be expensive.

The problem is that workload and workforce are no longer the same thing. A department may continue processing thousands of applications, returns, permits, claims, or registrations every month, but the amount of human effort needed to process them may fall sharply. A digital system can capture information automatically, verify identities, calculate amounts, identify missing documents, issue standard notices, and even flag unusual cases for review. 

The volume of work may remain high, while the human workload declines. That difference changes everything.

Changing face of cadre assessment

Cadre assessment can therefore no longer be only a headcount exercise. It must become a task assessment. 

Every government job contains different types of work. Some tasks are repetitive. Some require judgement. Some involve investigation. Some depend on empathy. Some require legal interpretation, negotiation, or discretion. Others simply involve moving information from one place to another. Technology will not affect all these tasks equally.

This leads to a better principle for future workforce planning: do not ask only whether a job will survive; ask which tasks inside the job will survive. 

A position may contain 10 important tasks. Four may be fully automated. Two may be partly automated. Four may continue to require human judgement. The position may not disappear, but it should certainly change. That is a much more realistic way to assess staffing needs than simply replacing every employee who retires.

A simple workforce equation may help: future human requirement = total work – automatable work + new human work. The final part is especially important because technology does not only remove work. It creates new work too. 

A digital revenue system may reduce data entry but increase the need for data analysts and risk specialists. An automated licensing platform may reduce counter work but create a stronger need for cybersecurity experts. AI may scan thousands of files, but experienced officers will still be needed to examine unusual cases, interpret complex rules, resolve disputes, and take responsibility for final decisions.

3 questions for cadre planning

The future may therefore require fewer people in some areas and more people in entirely new areas. That is why digitalisation should not become a simple staff-cutting exercise. The better question is not, ‘How many people can technology replace?’ It is, ‘Where should human capability be used when technology can do the routine work?’

One useful way to think about this is to divide government work into three broad categories: automate, augment, and human-centred. 

Routine, predictable, and rule-based tasks are strong candidates for automation. Data entry, standard calculations, basic document checks, and routine notifications fall into this category. 

A second group should be augmented rather than replaced. Auditing, risk analysis, forecasting, planning, and legal research are good examples. Technology can identify patterns and provide recommendations, but people should still interpret what those patterns mean. 

The third group includes work where human value remains strongest, such as handling sensitive complaints, negotiating complex cases, supporting vulnerable citizens, investigating unusual situations, and making decisions involving discretion.

Future cadre planning should therefore ask three simple questions. What should be automated? What should be augmented? What must remain human? These questions are far more useful than simply asking how many vacancies exist.

This also leads to an uncomfortable truth: a vacancy is not automatically a recruitment need. When someone retires, the first question should not be, ‘Who will replace this officer?’ It should be, ‘Does this work still need to be done in the same way?’ 

Can part of the process be automated? Can unnecessary steps be removed? Can two units share the function? Can the service be centralised? Can technology reduce manual intervention? Can an existing employee be reskilled? 

Recruitment should come after these questions, not before them. Otherwise, government risks hiring people today for work that technology may largely eliminate tomorrow. 

Public sector recruitment is not a short-term decision. A new employee can create salary, pension, training, and administrative commitments lasting decades. Poor cadre planning is therefore not just a human resource issue. It is a fiscal issue, a productivity issue, and increasingly, a technology governance issue.

The need for workforce intelligence

However, there is another danger. Governments can become too enthusiastic about automation. Public administration is not an online shop. Government deals with rights, law, fairness, vulnerability, and accountability. 

An algorithm may recommend an action, but someone must remain responsible. A chatbot may answer a routine question, but a distressed citizen may still need a real person. AI may identify a suspicious transaction, but suspicion should never automatically become guilt.

Efficiency matters, but so does humanity. The goal should therefore be less unnecessary human intervention, not less humanity.

This shift will also change the type of people governments recruit. Future departments may need fewer purely routine clerical roles and more people with skills in data analytics, cybersecurity, behavioural science, systems thinking, digital communication, and technology governance. 

Even traditional officers will need new capabilities. The most valuable public servant of tomorrow may not be the person who can move a file fastest. It may be the person who knows when the system is wrong, who can question an automated recommendation, who can combine data with judgement, and who understands that behind every digital transaction there may still be a human story.

Technology planning, workforce planning, and skills planning can therefore no longer operate separately. If a department expects major automation within five years, its recruitment decisions today should reflect that future. If certain skills are likely to become less important, reskilling should begin now. If new technical capabilities will become essential, governments should start building them before the shortage becomes a crisis.

This is why traditional cadre control may need to evolve into something more dynamic: workforce intelligence. Instead of focusing only on approved posts, grades, and vacancies, institutions should continuously examine workload, citizen demand, retirement patterns, employee skills, automation potential, productivity, and future services together. 

A modern department should be able to say: ‘This task is growing,’ ‘This process will soon be automated,’ ‘This skill is becoming less relevant,’ ‘This new capability will become critical,’ or ‘This vacancy should be redesigned rather than replaced.’ That is the difference between merely managing vacancies and actually planning the future.

Man and machine

Perhaps every new recruitment request should therefore face one simple test. Call it the empty chair test. Before filling a vacant post, ask: if this chair were still empty five years from now, would government truly need a person sitting in it?

If the answer is yes, recruit. If the answer is partly, redesign the job. If the answer is no, redesign the process.

The future of government should not be about protecting every old position. Nor should it be about replacing people simply because technology exists. It should be about using people where people matter most. 

Technology can carry repetitive work. Systems can handle the predictable. Machines can process the volume. People should carry what machines cannot easily carry: judgement, responsibility, creativity, empathy, and trust.

That may be the real answer to the future cadre question: not more people, not simply fewer people, but the right people, with the right skills, in the right roles, doing the work that genuinely still needs a human mind and a human touch.

(The writer is an independent researcher)

(The views and opinions expressed in this article are those of the writer and do not necessarily reflect the official position of this publication)