The Function That Cannot Be Neutral
The Function That Cannot Be Neutral.
What happens when a department is asked to be independent but doesn’t have the independence?
Here is a complaint you have heard, and may have made.
HR always sides with management.
It gets said quietly, after a grievance process that went nowhere, or a mediation that produced an outcome everyone could have predicted at the start. It is usually treated as unfair – a slur on a function that works hard at difficult work and often cares a great deal about doing it well.
I think it is neither unfair nor a judgement about anyone’s character. It is an accurate reading of where the function sits.
What mediation actually requires
Mediation is a defined practice with a defined condition. It requires a third party with no stake in the outcome. That is not a nicety; it is what makes the result credible to both sides. Remove it, and you do not have a compromised mediation. You have something that is not mediation at all.
Now consider what an HR function is typically holding while mediating between an employee and their manager.
It will usually have written the policy under which the dispute is assessed, and advised management on applying it before the dispute arose. It is now investigating. It will administer whatever consequence follows. In many organisations it will also help prepare the response if the matter reaches a tribunal.
How many of those sit in one place varies. But the accumulation is the norm rather than the exception, and each addition narrows the distance between the function and one of the parties. Add that its budget comes from the executive, its remit is set by the people it might have to find against, and it will be working with that manager on the next twelve cases, and the position is far from independent.
Which is why the employee’s assessment is correct before anyone has spoken. They are not judging a person’s fairness. They are reading a structural arrangement. A fact about the organisation, visible to both parties, unaffected by anyone’s intentions.
Two things this doesn’t say
It doesn’t say HR functions are partial. Most are scrupulous, and the individuals in them are frequently better at setting aside their own interests than the people criticising them. The point is that scrupulousness is unverifiable from outside, and therefore cannot do the work that position would do.
And it doesn’t say these functions are useless. HR resolves a great many disputes perfectly well, because most disputes do not turn on whose side the organisation is on. Two colleagues who cannot work together, a misunderstanding about a decision, a manager who handled something clumsily. None of those puts the organisation’s interest against the employee’s.
The argument applies to the subset where they genuinely diverge. And what makes that subset divergent is usually money. An upheld grievance is an admission; it creates exposure, invites a claim, and may establish a precedent across the workforce. A grievance not upheld creates none of those things. The organisation therefore has a direct and quantifiable interest in the outcome. And HR is the part of the organisation that carries that exposure.
That is precisely the subset in which independence was the point.
Why better practice cannot reach it
Almost everything written about this proposes that HR become more neutral: training, conscious attention to bias, techniques for hearing both sides, deliberate reflection before difficult conversations.
Much of it is sensible, and it improves the quality of the work. None of it addresses the problem, because independence is not a quality a function can demonstrate through its own conduct. It is a property of position, and the position was determined by whoever set the reporting line and the remit, which was not HR.
A department cannot establish its own independence. It can only be granted to it, and nobody has.
Where else this happens
HR is the most visible case, not the only one.
Internal audit is asked to report fearlessly on the organisation that employs it, to a committee appointed by the board it may need to criticise. The profession takes independence extremely seriously and has standards requiring it, and the function’s budget and headcount are still set inside the institution.
Compliance is asked to constrain commercial activity while being funded by it, and to refuse the people who determine its resourcing.
A whistleblowing function receives concerns about the organisation it belongs to, often sits within HR, and in many cases conducts the initial investigation itself.
A departmental sponsor team is asked to hold an arm’s-length body to account while sharing responsibility for that body’s performance. Also, it needs a working relationship with it indefinitely.
In each case an organisation has created a function requiring independence and placed it inside the thing it is meant to be independent of. Not through carelessness – genuine external independence is expensive and slow. The in-house version is cheap, fast, keeps the matter private, and satisfies the procedural requirement.
The cost, and who absorbs it
So the organisation obtains the form of independence at the price of a department it already employs. The arrangement looks efficient.
What it does is transfer a cost in two directions.
The employee or the concern-raiser pays the first part. They engage with a process described as impartial, discover it isn’t, and conclude the system isn’t for them. They will not use it again, and they will tell colleagues. The next person weighs that before deciding whether to raise anything at all.
The function absorbs most of the impact, and nobody designs for it.
HR carries the reputational consequence of a role it was never structurally able to perform. The distrust attaches to the function rather than to the arrangement, because the arrangement is invisible and the department is not. Every unbelieved process confirms the reputation. Every improvement in technique fails to shift it. And the function spends years being told it needs to build trust, when what it needs is a position from which trust would be possible.
That is accountability without authority in one of its purest forms. A department answers for an outcome – a process both parties believe in – that it has not been given the means to produce.
What would change it
Not better functions. Three structural things, in descending order of cost.
Genuine independence where the stakes warrant it.
External mediation, an external investigator for serious concerns, an audit function reporting to a committee with its own powers of appointment and removal. Expensive, so it should be reserved for cases where the cost of a disbelieved process exceeds the cost of buying a credible one. Organisations estimate that threshold badly, because the cost of disbelief arrives late and diffusely while the external fee arrives now and with an invoice.
Separation of the roles that cannot sit together.
The part of the organisation that investigates should not be the part that will later defend the position, and neither should be the part that advised on the decision being investigated. Splitting those, even internally, even imperfectly, removes the worst of it, and it is considerably cheaper than buying independence from outside.
And at minimum, be accurate about what is on offer.
A process described as impartial when it is not does more damage than one described plainly. “We will investigate this thoroughly, and we are the employer” is less attractive and considerably more trustworthy. People can work with a declared interest. What they cannot work with, is an interest denied.
The general point
Organisations routinely create functions that require a structural property, place them where that property is unavailable, and then treat the resulting failure as a performance problem.
The function is told to build trust. The training budget goes up. And the one thing that would change the outcome: where the function sits, what else it holds, and who sets its remit, is the element nobody touches, because touching it costs the organisation something real.
So the next time a process described as independent is not believed, the useful question is not whether the people running it were fair.
It is whether a department in that position could have produced a result anyone would believe.