… is from page 241 of the Mercatus Center’s 2016 re-issue of my late colleague Don Lavoie’s more-relevant-than-ever 1985 volume National Economic Planning: What Is Left?  (references deleted; link added; original emphasis):
When the central argument of this book, the knowledge problem, was first made  some sixty years ago, the radical idea of comprehensive planning as a hopeful alternative to the status quo was at its peak. Today the radical ideology of planning is intellectually bankrupt. All that remains are meek suggestions to try yet one more variation on the century’s dominant theme of noncomprehensive planning. But this policy does not resolve the knowledge problem; it merely substitutes a form of destructive parasitism on the market process in place of its earlier unachievable goal of dispensing with market processes altogether. The knowledge problem shows that the freely competitive market order makes more effective use of the information that lies dispersed throughout society than can any of its participants. This means that noncomprehensive planning is blind interference into a complex order, interference which can succeed in protecting and enhancing monopoly and privilege, but which cannot improve the productive capacity of a modern technologically advanced economy.
DBx: A frequently repeated fallacy today goes something like this: “Yes, yes, Mises and Hayek showed convincingly that full-on socialism – ‘comprehensive’ economic planning by the state – is unworkable. These economists demonstrated the inescapable reality that state officials can’t acquire and process all the knowledge necessary to consciously create an economic organization that performs as well as, and much less outperforms, the free-market order. But we advocates of industrial policy today aren’t proposing full-on socialism; we are proposing more modest interference with the market process. We don’t wish to totally replace markets with socialist planners; we wish only to replace parts of markets with democratically accountable officials who will help to steer the nation’s market to a better destination. We proponents of industrial policy today merely seek to prevent the free market from behaving like the drunk donkey that it is. We proponents of industrial policy today simply with to guide the market as a ship is guided with a compass.”
Sounds reasonable. But it’s not. While piecemeal use of tariffs here, subsidies there, and this prescription and that proscription issued by government bureaucrats invested with discretionary power are unlikely to ruin an economy as completely and as quickly as would full-on socialism, such industrial policy nevertheless cannot overcome the knowledge problem.
Contrary to the belief of today’s advocates of industrial policy, the knowledge problem as identified by Mises and Hayek does not kick in only with full-on socialism. The knowledge problem arises whenever politicians and their hirelings attempt to improve economic outcomes by overriding market processes. And this reality holds for politicians and their hirelings who are democratically accountable no less than it does for politicians and their hirelings who are as secure in their power as was a 15th-century Hapsburg.
The fact that, say, a protective tariff on some good deemed by government officials to be “economically strategic” will not alone crash an entire economy as would full-on socialism does not imply that that tariff imposes no net harm on the economy. The imposer of the tariff cannot know what he claims to know – namely, that the economic benefits to the country from the tariff will outweigh its costs. The imposer of the tariff cannot possibly understand the complexity of the economic forces at work  that cause the market price of the tariffed good to be what it is; he cannot possibly know what will be all the reactions of buyers and sellers, at home and abroad, to the price made artificially high. The imposer of the tariff – regardless of his academic credentials or his experience and success in private industry – can see only the surface of the economic phenomena into which he coercively imposes his visible hand. Yet beneath this surface lies vast, massive, incomprehensible complexity.
Just as shooting BBs from a BB-gun at a healthy person would not kill that person as would firing at that person several live rounds from an AK-47, industrial policy as proposed by people such as Marco Rubio, Elizabeth Warren, and Oren Cass would not utterly destroy an economy as would inflicting on that economy full-on socialism of the sort that was all the rage a century ago. But peppering a person with BBs still harms that person, and the heavier the peppering, the more harm the person suffers, even if he or she is never killed by the peppering. For the shooter of the BB gun to defend his actions by pointing out that he’s not executing his victim with an AK-47 hardly justifies the shooter’s interference with the person.
I’ve noted many times, and will continue to note, that advocates of industrial policy offer no explanation of how government officials would get the knowledge that they would need in order to direct the economy toward ‘outcomes’ that are better than those that arise in free markets – markets unimpeded with tariffs and subsidies. Advocates of industrial policy content themselves to repeat the banal fact that markets aren’t perfect, and then to gallop, rather like drunk donkeys, to the conclusion that therefore politicians and their hirelings who have the power to coercively override the market’s resource-allocation patterns will outperform markets. Just how politicians and their hirelings will achieve this happy outcome we are never, ever told.