Mises’ Human Action is an austere no frills explanation of not just how a market economy works, but also why only a market economy can work. There are the members of a community who have material desires they would like satisfied and personal services they would like to engage. Most of what individuals want is dependent on what has already been produced and sold in the past, although of those desires are for goods and services that have only just been made available.
There are also individuals who earn their own living by running businesses that produce these goods and services in the hope that others will buy them, and in so doing pay enough in total amongst all purchasers to cover the costs of production.
There are also entrepreneurs who run businesses that produce inputs that are used within other businesses. Ultimately, however, all production is focused on satisfying the demands of final consumers. It is in this sense that consumers call the shots. What people are willing to pay for determines what will be produced since only those enterprises that produce goods and services that earn a profit can stay in business.
But what people will be willing to pay for is an unknown that can only be discovered if an entrepreneur makes the decision to produce some good or service and put it on the market. Only then can it be discovered whether whatever has been produced can be sold at a profit. Once it has been determined that a profitable enterprise can be established to produce these particular goods and services, many other firms may then follow along and try to produce the same product or even better versions.
This is how the market works through the trial and error efforts of entrepreneurs to find products that can be sold at a profit. If a community is content never to change any of the products it chooses to buy, and there are never any interruptions or changes in the supply conditions for the inputs used in production, the economy can enter a steady state which can repeat endlessly the same routine. But since in the real world there are new innovations taking place all the time, and changes in the supply conditions for inputs, a steady state outcome is an impossibility.
Therefore to ensure an economy continually improves the products produced, and can adjust to new conditions in the supply of inputs, a market mechanism is essential. No other mechanism will work if a community is intent on improving its standard of living or wishes to accommodate changes in the conditions of supply.
The question then is whether there is any role for government oversight and regulation in such an economy. And while it is clear that Mises is reluctant to state that there is such a role for governments because of the principle of give-them-an-inch-and-they-will-take-a-mile, nevertheless, he does accept that government does indeed have such a role. This is from Human Action:
There are certainly cases in which people may consider definite restrictive measures as justified. Regulations concerning fire prevention are restrictive and raise the cost of production. But the curtailment of total output they bring about is the price to be paid for avoidance of greater disaster. The decision about each restrictive measure is to be made on the ground of meticulous weighing of the costs to be incurred and the prize to be obtained. No reasonable man could possibly question this rule. (Mises  1963: 748)
There ought to be no doubt from this passage that there are circumstances for which government regulation is warranted. It is a cost-benefit calculation in which regulations are laid down, which have a cost in lost production, but in which there is a positive return in the prevention of an even more costly outcome whose probability of occurrence has been reduced.
His reluctance to state in a more fulsome way that such regulations have a role in economic management is based on his no doubt correct judgement that from the example of this unquestionable use of government regulations to diminish the possibility of a much more costly outcome has been a thin edge of the wedge to justify an enormous and monstrous regulatory regime across all the economies of the world. If anything has occurred, the meticulous examinations that now occur are to determine if there is absolutely no possible harm that might occur if some regulation is not introduced and enforced. The weight of evidence has now been placed on those who wish to reduce such regulations where outcomes with a small probability of occurrence are not made the basis for such rules.
The principle should therefore not be seen as a blanket ban on government regulation per se, but as the need for those who wish to impose such regulations to demonstrate that the potential risk is large and that the market would not be expected to provide its own cure if left on its own to work things out.
If, for example, individuals who wish to built houses in the middle of flood planes that are expected to flood only once in fifty years should be permitted to do so, but also told that if they do, they must cover the cost of insurance themselves and not expect a government to make good any losses they might endure because of flood damage.
And while there is need for licensing for doctors and electricians, since no consumer can be expected to research into the competencies of individuals who declare themselves a doctor or electrician, there is no need for regulation in endless other occupations, with hairdressers as the most notorious example of regulatory overkill.
But there is a further issue in relation to regulation. Within political debate, to argue that no regulations are ever justified will instantaneously lose the public debate. No one will accept that the market can be left to itself without oversight and regulation. Finding the balance is important, but not to recognise an important social function of regulation by governments, specially by governments under popular control, is to throw the baby of good economic management out with the bathwater of heavy-handed control.
Mises, Ludwig von.  1963. Human Action: A Treatise on Economics. Fourth Revised Edition. San Francisco: Fox and Wilkes.