The Roots of the Left-Right Divide: Whose Suffering? And Who Knows Enough to Act?

The Long Read

In January 2023, Jacinda Ardern resigned as New Zealand’s Prime Minister after five years in office. She left as one of the most celebrated progressive leaders of her generation – and as one of the most domestically repudiated. Labour’s vote virtually halved between her historic 2020 majority and the 2023 election, and the party polled higher in the weeks after she resigned than it did while she led it. She now holds fellowships at Harvard and Oxford, commands global audiences, and has written a New York Times bestseller. She did not tour New Zealand for the launch.

The gap between global icon and domestic reckoning is not an anomaly. It is a compressed version of the left-right divide made visible in a single career – and the divide is not unique to New Zealand.

Ardern was the most explicit recent practitioner of progressive governing philosophy taken to its full extension. She declared “kindness” the operating principle of her government. During the pandemic, she told New Zealanders: “We will be your single source of truth. Unless you hear it from us, it is not the truth.” The line captured more than Ardern may have understood: a habit of treating the state as the proper arbiter of what voters should be permitted to believe.

Kindness and the source of truth were not slogans. They were the moral and epistemic commitments of the modern left, stated plainly as governing doctrine by a leader with the mandate to act on them.

Ardern was making two claims at once.

The first was moral: whose suffering matters most, and when? The left’s invariable answer is the suffering you can see now. The right’s focus is typically the suffering that comes later, transmitted through the incentives today’s decisions set in motion.

The second claim was epistemic: who can be trusted to know what to do about it? The left’s answer is typically that central authorities, equipped with enough expertise and data, can design solutions to society’s problems. The right’s answer is that no authority can know enough – that the knowledge required to make human lives flourish is dispersed across millions of individuals, embedded in local circumstances that the centre cannot see.

Each side has an answer to each question, and the answers cluster. Those who weight present suffering most heavily are almost always the same people who trust government to relieve it. Those who weight future harms most heavily are almost always the same people who doubt anyone knows enough to act. The clustering is what makes left and right look like coherent worldviews rather than positions on two separate dimensions. It is also what makes each tribe genuinely unintelligible to the other.

You can see the clustering in practice. From Sunstein to Ocasio-Cortez – and, in New Zealand, Ardern – the progressive cluster holds: present suffering weighed first, government trusted to act. From Hayek to Thatcher – and, in New Zealand, Ruth Richardson – the opposing cluster holds in mirror: future harms weighed first, government distrusted by default. The positions travel together.

The moral question

Jonathan Haidt’s work in The Righteous Mind identified something that changes how political argument looks once you see it. Haidt showed that moral judgement runs on a small set of intuitive foundations – care, fairness, loyalty, authority, sanctity, and liberty – and that left and right respond to these foundations in systematically different patterns.

The left reasons primarily from care and fairness, and harm to identifiable individuals triggers a powerful moral response – immediate and vivid and hard to argue against. The conservative right draws on a wider palette: proportionality, loyalty, sanctity, and respect for norms that have evolved, slowly and painfully, to channel human behaviour in sustainable directions. Both palettes include genuine compassion. They simply have a different focus.

The left’s moral compass points toward the person suffering in front of us. The child going hungry. The patient denied a treatment that would have saved them. The worker ground down by a powerful employer. To tolerate visible suffering when relief is available is cruelty.

The right’s moral compass points toward the person who will be harmed by the incentives today’s decisions set in motion. The future children born into dependency because the system made it costless. The future target of crimes that weakened deterrence failed to prevent. The next generation inheriting institutions hollowed out by good intentions. To ignore future suffering because it has not yet arrived is also cruelty.

Both sets of people are real, and both sets of harms are real. What differs is not compassion but time.

Ardern’s pandemic response was the moral compass of the modern left in operation. The visible suffering – Covid deaths, vulnerable people exposed, frontline workers under strain – was weighted heavily. The costs of the response – children’s lost schooling, businesses destroyed, families separated, the cancers not diagnosed because screening had stopped – were weighted lightly, because the people bearing them were as yet unknown. This was not a failure of compassion. It was what the moral axis looks like when one side of the timeline is visible and the other is not. Decent people, weighing what they could see, choosing to relieve the suffering in front of them at the price of suffering they could not see and would not have to look at.

The right’s critique of that posture is not that the present suffering did not matter. It is that the suffering not yet visible mattered too, and was not weighed. That critique sounds like callousness only if you assume the timeline ends with the here and now.

The epistemic question

The moral compass is half the story. The second commitment, which travels with it so reliably that most people experience the two as a single position, is about who is trusted to know what to do.

The left’s answer runs as follows. Social and economic problems can be addressed by the right combination of planning and authority. Industrial strategy, optimal price levels, the allocation of housing, the structure of competitive markets – these are not mysteries but problems for expert design, identified, measured, and acted on through policy. When a policy fails, the lesson is to give the experts more resources, not to question whether the approach itself is sound. Dissent from expert consensus is not principled scepticism; it is ignorance, bad faith, or the special pleading of those whose interests are threatened.

The right rejects the left’s premise. No central authority can know enough to design solutions for the lives of millions of people. The information that matters – what they need, what they value, how their circumstances are changing – is held by those people themselves, in forms that cannot be written down or aggregated. The further a decision sits from the people it affects, the less likely it is to fit their actual circumstances. The right’s confidence is not in the absence of expertise. It is in the people closest to the problem, acting on knowledge the centre cannot reach.

Nobel laureate Friedrich Hayek named this the knowledge problem, one of the most important and underappreciated ideas in the social sciences. Hayek’s case was historical as much as theoretical. The Soviet grocery shelves of the late twentieth century were empty not because the planners were bad people, or because they lacked information about factories and farms. They were empty because the price system the planners had replaced was the mechanism that turned dispersed knowledge into coordinated action. Without it, no amount of central data could close the gap.

Not every progressive thinker dismisses the knowledge problem. Some of the most important work on how communities manage shared resources comes from the left, and it takes the limits of central planning seriously. Elinor Ostrom won the Nobel Prize for showing that local communities – Swiss villages, Spanish irrigation cooperatives, Maine fishing fleets – can govern shared resources sustainably through their own evolved rules, without either markets or central direction.

But Ostrom’s model was not what New Zealand built into its pandemic response, and it is not what typically shapes welfare and social housing systems. Ostrom’s model also requires social conditions – close, continuous, self-policing communities – that modern democracies do not have at scale. The politics of visible suffering rewards what governments can point at: a government that delegates decision-making does not get to point at the relief it has provided. The moral axis selects against the more sophisticated epistemic position because that position is harder to praise.

New Zealanders do not need to reach back to the Soviet Union to see the knowledge problem at work. They lived through it five years ago. Managed Isolation and Quarantine was designed by people who cared, staffed by people who cared, and approved by a Prime Minister who believed, sincerely, that a well-run central system could protect New Zealanders from a dangerous virus while honouring the rights of citizens to return home.

Charlotte Bellis was a New Zealand journalist working for Al Jazeera in Qatar when she discovered, in the middle of 2021, that she was pregnant. Qatar criminalises pregnancy outside marriage. She flew to Afghanistan, where she had contacts from her reporting on the Taliban takeover, and applied for a place in MIQ so she could come home. She was refused. The emergency allocation system the government had built could not find her a place inside the criteria its designers had written. After her story was published, senior Taliban officials, in a detail it is hard to make up, offered her their personal protection if she chose to stay in Kabul.

A pregnant New Zealand citizen had been told, in effect, that she would be safer under the Taliban than returning to her own country. The public outrage was loud enough that the government eventually granted her an emergency voucher. The designers had not changed their minds. They had simply been reached. The High Court eventually ruled that the system had operated unjustly and infringed New Zealanders’ right to return home.

Multiply that case by the thousands that never reached a front page. A father who missed his daughter’s funeral. A dying cancer patient who could not get home for treatment. Each was judged, under the emergency allocation framework, as falling outside the criteria the designers had written. The designers were not cruel. They simply could not see what they had not thought to measure, and what they had not thought to measure was a good deal of what mattered.

A confident system, built by decent people, on the premise that a central allocator could weigh millions of circumstances it could not see. This is the knowledge problem made flesh.

Housing policy tells a quieter version of the same story. Auckland’s planners have long specified balcony requirements, glazing ratios, and parking spaces on apartments priced beyond the reach of first-home buyers. Each requirement sounds reasonable in isolation. Together, they add tens of thousands of dollars to the cost. The buyers would trade these features for a lower price. They are not asked. The planners know what good housing looks like. The price of that knowing is a housing affordability crisis.

The shape is the same in each case. The Soviet shelves were empty for want of information. MIQ substituted judgement for the circumstances it could not see. Auckland’s planners substitute their preferences for the buyers’ own. In each case, central design overrides the mechanism that would have generated the knowledge the designers needed.

Of course, there are narrow cases where centralised action is justified. Defending the country. Enforcing contracts. Building roads. Stopping pollution that crosses property lines. The argument is not that government never knows enough to act. It is that government rarely knows as much as it assumes when it acts.

Beyond Sowell: what the two-question frame adds

In A Conflict of Visions, American economist and social theorist Thomas Sowell traced the political arguments of the previous two centuries to a deeper disagreement about human nature itself. Sowell’s “unconstrained” view sees human nature as malleable and social problems as tractable to reason and design. His “constrained” view sees human nature as fixed within narrow limits and social arrangements as the slow accumulation of trial and error, sustained by institutions whose value is not always visible to those who would reform them. From these two starting points, Sowell argued, the recognisable shapes of left and right follow: one yielding the politics of redesign, the other the politics of inheritance.

The two-question frame addresses the same divide that Sowell mapped, but it asks different questions about it. Sowell’s distinction is anthropological. It tells us what each side believes about human nature, and from that belief, predicts how each side will argue. The constrained vision carries an epistemic claim of its own – the limits of reason are part of what it constrains – but in Sowell’s account the moral and epistemic positions are derivative of the deeper anthropological priors.

This essay’s two-question frame is analytical. It does not ask what each side believes about humanity. It treats the moral and epistemic questions as logically independent, and asks what each side is doing in argument: which question each is answering, and why the answers cluster as they do.

That difference yields something Sowell’s frame cannot reach.

His account predicts that left and right will hold different positions across the domains of politics. The two-question frame predicts the same. But Sowell’s frame cannot explain why left and right experience each other not merely as wrong but as unintelligible. By treating the cluster as an expression of one underlying anthropological disagreement, the constrained-unconstrained frame implies that the two sides should hear each other as advancing rival answers to the same question. They do not. They hear each other as evading.

The two-question frame explains why. Each side is answering a different question, and the question the other side is answering is invisible to it. When the left makes a moral argument – this child is suffering now – the right’s response sounds like evasion, because it answers an epistemic question (no one knows enough to engineer the outcome) without engaging the moral one. When the right makes an epistemic argument, the left’s response sounds like evasion for the same reason. Neither side is reasoning badly. Both are answering different questions, and neither has noticed.

This is the puzzle Sowell’s frame leaves intact. The two-question frame solves it. And the solution matters, because it reframes what political argument across the divide is for. If the two sides are simply rival anthropologies, the argument is a contest of worldviews, won by whichever vision is closer to the truth about human nature. If the two sides are answering different questions, the argument is something else: a sustained mutual misunderstanding that no amount of evidence on either question alone can resolve. The first picture invites each side to defeat the other. The second invites each side to ask what the other is seeing that it is not.

The pattern in welfare

The two-question framework illuminates disputes that would otherwise seem to share no common structure. Take the argument that has shaped New Zealand politics more than any other over the past decade.

Welfare debates turn, on the surface, on how generous benefits should be and how strictly they should be conditioned. Underneath, they turn on something deeper. The left looks at a sole parent raising children on a limited income and sees hardship the state can relieve. A modest increase in the benefit, combined with wraparound support, will reduce child poverty and improve outcomes for the next generation. This is not an unreasonable expectation. Children living in material hardship do worse on almost every measurable dimension of life, and the intuition that giving the family more money will help is grounded in something true.

In New Zealand, a sole parent with one or more children receives around $500 a week in core support, supplemented by accommodation assistance, child support, and Working for Families – an income structure the left reads as inadequate to lift children out of poverty and the right reads as sufficient to support family formation without a partner.

The right looks at the same family and sees something different. Not the suffering in front of it, but the suffering the system’s incentives will produce among families not yet formed. If a benefit sufficient to raise children is available without conditions, household decisions about family formation and partner stability shift at the margin. Over a generation, the patterns of family and work that sustain communities erode, and the children born into those patterns carry the costs. The right is not indifferent to the child in front of it. It is weighing the child not yet born, whose life will be shaped by what today’s choices set in motion.

Welfare is one domain among many. The same two questions run through others – education, criminal justice, the regulation of speech – and produce the same pattern. The moral axis is what we have shown here. The epistemic axis runs through welfare too: any rules-based system produces outcomes the rules cannot see, and the case manager with ninety minutes and a form has a life the form does not reach.

Both axes hold across the domain. Each side weighs the suffering it can see and trusts the institution it was inclined to trust. Each side then accuses the other of indifference to the suffering it does not weigh, and arrogance about the institution it does not trust. The cluster holds across the domains.

The return to Ardern

When New Zealand turned against Ardern, it was not repudiating kindness. It was naming, imperfectly and without the vocabulary, the costs it had paid for a style of government that treated the unseen as invisible and the unknowable as known.

The pattern is not specific to her or to New Zealand. Democracies across the developed world have been governed, over the past decade, by leaders operating on the same cluster: present suffering weighed first, central authorities trusted to act, what could not be measured weighted lightly because no one had to look at it. Each country has had its own reckoning. Each reckoning has been described in its own vocabulary, by people who lived through it and who retain the right to say what they lived through.

The argument worth having begins when each side stops accusing the other of moral failure and starts asking which question the other is answering.

This essay is part of an ongoing series on liberalism, democracy, and the international order. Related writing in Persuasion, Quadrant, Quillette and on Plain Thinking is collected here.