Isn’t war massive economic stimulus and didn’t World War Two finally drag the United States out of the Great Depression? Why not a government to mandate a maximum life on all automobiles? Why doesn’t the government mandate a certain minimum “fair” compensation for workers? Wouldn’t these workers then be able to spend more money? Wouldn’t that be good for them and for the economy? Destroy the old, rebuild with new, raise wages for everyone. Oh the growth, the jobs, the wealth! 

Snarkiness aside, one of those ideas was actually enacted in 2009 as “Cash for Clunkers”, only to see a National Bureau of Economic Research (NBER) working paper published finding “no evidence on employment, house prices, or household default rates” benefits from the $3 Billion program. Another subsequent NBER paper concluded that sixty percent of the subsidies went to households that would have purchased during the period anyway. The remainder did accelerate auto sales by less than eight months. However, this initial acceleration was matched by an equal reduction in sales in subsequent months. The program’s cost in hype, inefficiency, and distraction went unmeasured. 

Economic quackery is often passed out in soundbites that appeal by highlighting obvious benefits of a proposed action, while ignoring sometimes obscure or delayed negative effects. This is particularly effective when the positive benefits are concentrated on a certain economic group (whose members lobby for the action of course), and the negative effects are widely dispersed across a larger population.

"There is a persistent tendency of [people] to see only the immediate effects of a given policy, or its effects only on a [certain] group, and to neglect to inquire what the long-run effects of that policy will be not only on that group, but on all groups. It is the fallacy of overlooking secondary consequences. The art of economics consists in looking not merely at the immediate but at the longer effects of any act or policy; it consists in tracing the consequences of that policy no merely for one group but for all groups" -Henry Hazlitt

Many of ideas and examples in this piece are borrowed from portions of Henry Hazlitt’s wonderful tome “Economics in One Lesson”. Originally published in 1946 and most recently updated in 1979, it combines cogent readability with biting wit. Hazlitt displays an interesting and often humorous perspective on economics, liberally and pointedly laying all about with his version of Occam’s Razor. 

The Broken Window

One of our favorite examples of secondary effects and their importance is Hazlitt’s example of a broken window, and the economic consequences that follow. Let’s assume a young vandal has just thrown a brick through a grocery store window. The store manager runs our furious but the perpetrator is gone. A crowd gathers and begins to stare at the gaping hole in the window and the shattered glass. 

After a while, the crowd feels the need for philosophic reflection on the act and its consequences, and several people are almost certain to remind the store manager that there is a bright side to the misfortune. After all, it will create some business for a glass maker. How much would such a window cost? Maybe a few thousand dollars? That would mean a decent amount of business for the glass-maker, but the possible benefits wouldn’t stop there. In fact, the glass maker will pay a salary to the installation technician who will then be able to spend money on goods and services. This will in turn bring additional business to those merchants. So the smashed window will go on providing money and employment in ever-widening circles! The conclusion of the crowd is that the destructive act was actually a publicly beneficial one.

The crowd is in fact correct in their initial conclusion that the act of vandalism will mean additional business for the glass manufacturer and another job for the window installer. However, the store manager will be out of the few thousand dollars he planned to spend on a new delivery vehicle (or some other equivalent need). Instead of having a window and a delivery vehicle that day, he just has the new window and no vehicle. If we think about the overall economic community as including the glass maker, the auto dealer, and the grocery store owner, there is no net gain and in fact there is a loss because the resources that would have “produced” the delivery vehicle merely “replaced” the window.

The key here, is that by forgetting about what the store manager could have bought with the money he instead paid to replace the window, the crowd’s calculus only considered roughly half of the economic consequences of the event. This is unfortunately common in economic analysis. It is easy to forget entirely about the negative consequences of actions and transactions that never happens. These obscured economic opportunities that failed to become activities, are what is meant by “secondary consequences”.

The False "Benefit" of Destruction

Having finished with the Broken Window, let’s apply this same line of thinking to destruction, war, and ultimately Cash for Clunkers. The key thought process remains the same, specifically the consideration of the secondary consequences. In the case of the destruction caused by war (or any other physical calamity), and the subsequent increased demand often cited as an economic benefit, it is helpful to think in terms of property and of the exchange of value rather than the more abstract notions of supply, demand, money, and GDP. No one would want to have his own property destroyed by a war. Anything that is harmful to an individual cannot be beneficial to a collection of individuals in net. 

Of course it is true that destruction can move benefits from one pocket to another (from the auto dealer to the glass manufacturer in the Broken Window example). In the case of the World War Two, it was indeed destructive to the global economy overall and Europe specifically, but proved beneficial for some period to the American economy while the United States remained out of harm’s way. However, it is important that this shifting in resources results in a net loss to the “system” (however defined) even while creating relative winner and losers within the “system”. The magnitude of destruction and loss evident in Europe and Asia went far beyond the beneficial economic growth that happened in the United States.

This is exactly what happened with the Cash for Clunkers program, initially pitched as a net economic stimulus. In the short run, it is easy to list the winners. Automakers certainly weren’t hurt. At worst they sold more vehicles now and fewer later. The most obvious beneficiaries on the consumer sider were the fortunate few who were planning on buying a new automobile to replace an aging vehicle anyway. Free money fell in their lap. To a lesser extent, as the NBER research concluded, benefits accrued to those planning to buy a new vehicle in the next eight to ten months. These individuals were able to gain marginally by “pulling forward” their vehicle demand (in exchange for a rebate) by a few months. Of course, this meant that the demand that otherwise would have been there in the future went missing-in-action.

But government cannot create wealth (barring certain special circumstances). What was given to the beneficiaries of the Cash for Clunkers program had to be taken from someone else in the form of taxation. What would these funds have purchased? What investments or consumption would have been improved? No one will ever know for sure, and these secondary consequences may be estimated, but will never be measured.

Now of course, critics of this line of reasoning would point out that the new automobiles are more efficient and therefore our analysis misses the benefit of the upgrade. Similarly, it was sometimes said that post-war Germany and Japan had an advantage over their American industrial rivals because they had newer plants and equipment (due to the need for replacing destroyed factories). But if this were really an advantage, the Americans could have offset it immediately by wrecking their old plants and junking the all of the old equipment. This didn’t happen. Owners of truly worthless Clunkers could have towed their vehicles to junkyards and bought new already, but they hadn’t until they were offered the hefty rebate.

There is an optimum rate of replacement for old plants, equipment, and automobiles. This optimal replacement occurs when new utility is significantly enough improved over old utility to justify the investment (given utilization and interest rates). The only advantage in destruction for a manufacturer would be if she were to have her plants destroyed at exactly the moment of optimal replacement. At any other time, destruction of old plants and equipment is merely destruction of capital. Destruction of capital, however marginal it may be, is never advantageous to the owner of that capital.

Minimum Wages

While we’re at it, let’s wade directly into the emotional and often politicized topic of wage controls (i.e. minimum wage). It seems silly to think that general prosperity could be increased by artificially boosting prices doesn’t it? Similarly, very few would today argue that general government price controls (mandating low prices) lower was doing anything but harm to its economy. To that effect we have the shining example of Soviet Union. But surprisingly many people have a concept that a minimum wage is somehow economically beneficial. If it isn’t true with prices, it can’t be true with wages. In fact, the opposite is true. The more such a law attempts to raise wages, the more harm does. 

To flesh this out a bit, let’s say that government mandated a minimum full-time wage equivalent to $100,000 per year. The first thing that would happen is that no one who is not worth $100,000 per year would be employed at all. You can’t make an employee worth a certain sum by making it illegal to offer him or her anything less. Such a policy simply substitutes unemployment for low wages. Under such a mandate, the former employee capable of $75,000 worth of economic output is now claiming unemployment because no company can afford to take incremental losses for each employee. 

There is an economic concept of equilibrium. Equilibrium wages and prices are those that equalize the production of stuff and services (supply) with the consumption of that same stuff and services (demand). If wages are pushed higher than the equilibrium level set by market forces in the employment market, the effect would simply be to reduce the supply of jobs and therefore increase unemployment. The best level for wages is not the highest level, but rather the level that results in the best possible levels of overall production and employment.

To add two minor disclaimers, there are a (very) few special situations where wages could be held below “market value” and in such specific cases government action may be appropriate. Also, these comments are not political in nature. We certainly aren’t “against” wage earners or high wages. However, we must consistently point out the economic costs of secondary consequences. If improving the lot of a particular set of workers was the policy goal (and a certain level of associated cost considered acceptable) then there are more economically efficient ways to accomplish that goal than falsely justified wage and price controls.


We began with some questions (with acknowledged sarcasm). But each of those demonstrably incorrect positions, have been taken by various experts and talking heads. These type of fallacious economic thinking do real economic damage to everyone in the economy. Interference in normal market mechanisms no matter how well-intended, has adverse secondary consequences even when those effects are difficult to specifically point to and measure. 

The Broken Window initially appeared to be net positive to the local economy. Destruction wrought by war or natural calamity is often misconstrued as beneficial. A minimum wage sounds like a benevolent policy. Until we trace the full consequences over time, the soundbite quackery can be appealing. But we must account not only for the immediate impact of a given economic action or policy on a particular group, but also the impact on all groups over all time. Secondary consequences (things that don’t happen) are tough to measure and are often overlooked. 

Cash for Clunkers was a real government program that passed Congress with little resistance. The concept sounded good, and the soundbites cheery. What could possibly be wrong with replacing old clunkers with newer more efficient vehicles? But secondary consequences were swept under the rug or ignored altogether. There is simply no free lunch in economics. Benefits and behavioral incentives can be directed to particular groups, but there is always a cost for doing so. In many cases the costs far outweigh the benefits.


Hoekstra, Puller, & West, “Cash for Corollas: When Stimulus Reduces Spending”, NBER Working Paper No. 20349, July 2014

Mian & Sufi, “The Effects of Fiscal Stimulus: Evidence from the 2009 ‘Cash for Clunkers’ Program”, NBER Working Paper 16351, September 2010

Hazlitt, Henry, “Economics in One Lesson”, Three Rivers Press, NY, New York, 1979