Under the right conditions, competitive, free markets succeed. (In this case the word `succeed' means that competitive, free markets efficiently allocate economic resources; there is no claim that they provide security or justice or other non-economic benefit. Non-economic benefits are not an issue in this discussion.) For competitive, free markets to succeed, conditions are key.
Among others, three conditions must be met:
Of course, we know that these conditions fail: people do not know everything. Cars and other `goods' release exhausts, which are external `bads'. Moreover, steel and flour mills, oil refineries, railroads, radio broadcast systems, and automobile manufacturing are examples of century-old industries that have high initial and low incremental costs.
(It cost Henry Ford a great deal of money to build his Rouge River plant, but once built, it cost relative little to manufacture an additional 100 Ford cars each year, up to a maximum.)
Products that are dependent on information, such as medicines, and products that are pure information, such as songs and software, are examples of current goods with high initial and low incremental costs.
A government can permit a market to allocate goods when people know of risks, when private and social benefits are the same, and when no industries with decreasing costs exist.
But when any of these features are missing — when investors do not need a modern corporation with limited liability, when they do not need bankruptcy laws, when there are no negative externalities like thrown-away paper, when the steel, automobile, and software industries vanish — then governments have a job.
There are reasons for governments to regulate economies. And, in theory, governments can do the job, or at least enough of the job to help a little.
However, in practice, governments often fail. The people in governments act to promote their interests, or the interests of their associates, rather than the interests of their country. Thus, in the latter 19th century in the US, railroad companies used the Interstate Commerce Commission to prevent competition among themselves that they felt was dangerous. In the 20th century, major US food companies `captured' the US agency set up to regulate them. While the food they sold became safer, at the same time, they reduced market competition against themselves.
So the issue becomes one of governance: what institutions will enable you, a citizen, in conjunction with other citizens, make sure that your agents do as you wish?
This is a traditional `agent/principal' question, except that it is applied between politicians and citizens to rather than between employees and their managements or between civil servants and politicians who are in office.
You and others citizens are the `principals': you give the orders. In theory, your `agent' acts on your behalf.
For example, you may not know why you are feeling ill; but an agent might: in this case, he would be a medical doctor. So you `go to the doctor'. He knows more than you about medicine.
(Indeed, the medical market is enabled by a lack of information on your part and the existence of specialized information on the part of others. If you could treat yourself, you would not need to visit a doctor. Similarly, you do not know about your future health or accidents. Consequently, in the US, many people who can afford it purchase health insurance.)
If, in your opinion as a `principal', your doctor, your `agent', fails to do his job, you switch to another. If you like him, you continue to visit him. By your actions, you provide your agent with information telling him whether his actions are perceived as beneficial to you, the principal.
When you cannot switch — perhaps your doctor is the only doctor in town — or when you do not know enough to decide when to switch, your actions as a principal will fail.
Then your so-called agent will be free to do as he or she likes. He can shirk. She can maximize her income. He can enhance some other personal goal. For example (to talk about a problem a friend of mine, a nurse, just mentioned), she can help a large company convert a fatal condition to a chronic condition that can be maintained through continuing treatment rather than find a cure that implies a one-time treatment. While saving lives is good, the social cost (and private cost to you) of suffering a chronic condition is worse than the benefit, both public and private, of a cure. But the cost to you and to the public may, depending on institutional motivations, be profitable to some.
As the Nobel Prize winning economist, Douglass C. North wrote1
... institutions basically alter the price individuals pay...
Moreover, it turns out that details matter: if citizens do not learn about the failings of their agents, they will not vote them out of office. This means that citizens, or their other agents such as journalists, must not only pay attention, they must not be cowed. Your other agents must tell you what is really going on. If they are cowed in any way, or if you are cowed, they or you will be poodles, not tigers.
There are more details. As Adam Przeworski says in the essay which inspired this Web page, A Better Democracy, A Better Economy,
What's needed ... is a clear party system with stable parties, a vigorous opposition, an effective system of checks and balances, a decent level of information that focuses on general economic performance, and non-electoral mechanisms for control over specific policy realms or particular organs of the government.
Without these features, neither a market nor a command economy will be efficient at allocating economic resources. And without some degree of economic efficiency, neither you nor anyone else will be able to afford security, justice, or beauty.
In the present world, for example, the current population is too large to be supported by the old technologies of the past. We could, if we chose, now feed everyone on the planet. But we could not even think of doing that if we were limited to the economic efficiency of a century ago.