Workers are the most visible losers in a recession, and on most measures they are also the heaviest, but the claim needs qualification before it can be accepted.
The case that workers bear the most rests on the nature and permanence of what they lose. Unemployment removes a household's entire income at once and replaces it with benefits worth a fraction of a wage, and a household with no savings has no way to absorb that. The damage does not end with the recession: a worker unemployed for a long period loses skills and confidence, employers treat a long gap as a signal, and studies of earnings consistently find that people who lose jobs in a downturn earn less for many years afterwards than otherwise identical people who did not. There are also costs that never appear in the statistics — ill health, family breakdown, children's education disrupted by a move or by poverty. Workers are additionally the least able to protect themselves: they cannot diversify, because they have one job and one income, whereas a shareholder can hold many different assets.
The case that others bear at least as much is that the losses elsewhere can be absolute rather than temporary. A firm that goes out of business is not recovered when the economy recovers — the owner's capital is gone, and for someone whose life savings were in the business the loss may be permanent in a way that unemployment is not. Shareholders lose value, and pensioners whose income depends on those investments may see a permanent reduction. The government faces falling revenue and rising spending simultaneously, and the resulting debt is repaid by taxpayers for decades. Consumers who keep their jobs face uncertainty and postponed plans. And a recession that reduces investment lowers the economy's capacity, so the eventual cost is borne by everybody, including people who were children at the time.
My judgement is that workers bear the heaviest cost, but only certain workers, and that the distribution within the group matters more than the comparison between groups. A recession does not reduce every worker's income a little; it removes a minority's income entirely while leaving the majority employed and, in some respects, better off — those who keep their jobs face lower interest rates and weaker price rises. The burden is therefore extremely concentrated, and it concentrates on the least secure: the young, those on temporary contracts, those in the industries most exposed to a fall in demand, and those with the fewest qualifications to find something else.
That concentration is what makes the comparison with other groups misleading. A shareholder whose portfolio falls 30% is poorer; a worker who loses their only job has no income at all. The proportion of what each stands to lose is quite different, and it is proportion rather than amount that determines hardship.
It also explains why the policy response matters so much. If the burden fell evenly it would be a matter for general stimulus alone; because it falls on an identifiable minority, unemployment benefit, retraining and targeted job support do more to reduce the human cost than the size of the stimulus by itself.