The Eyes in the Swiss Cheese
Take Swiss Cheese, the variety that comes with holes (technically known as eyes). There is a clear correlation between holes and cheese: the more cheese there is, the more holes there are. If you purchase twice as much cheese, you'll get a volume of holes twice as big. Since correlations are symmetrical, this also means the more holes there are, the more cheese there is. If you purchase enough cheese to double the volume of holes, then you have purchased twice as much cheese. Thus, a static mind satisfied with correlations may conclude that a good way to increase the total quantity of cheese is to increase the total volume of holes - which may be achieved quite simply by drilling holes in the given supply of cheese. Of course, this means fails, because it changes the proportion of holes to cheese, whereas the measured correlation upon which the reasoning stands crucially depends on this proportion or multiplier being a constant. Yet that's exactly how macroeconomic regulation by government works: find some existing correlation between some kind of wasteful government spending and a measure of general welfare, and then forcefully increase the spending in the hope to increase welfare...
The way that economic intervention works in modern social-democracies, State-sponsored econometrists gather lots of statistics, and make up arbitrary mathematical models based on correlations between said statistics. They are proud of their method that they call scientific because it uses mathematical tools from the physical sciences; only their method misses the one most crucial point of science: being a faithful description of the world. Just because they claim to observe the world while using numbers makes them no more scientific than are astrologists and numerologists. Since only hand-picked correlations are ever measured while causation is left to the arbitrary conjectures of the so-called economists, said economists, based on the right choice of statistics, can make up models with an arbitrarily malleable conclusion in terms of recommended government policy. Of course, so that the arbitrary choice remain in their hands, power-seekers will only sanction those models that confirm their plans, as made by their allies, lackeys and other useful idiots. The State will select which amenable appointees are anointed with the magic power to bless models as official, while political lobbies and mass media propagandists will work hard to put forward those models that support the policies they favor. And so, justified by some or some other fake models, governments intervene.
Now, once chosen their official models built out of correlations, Politicians will purport to improve the economy by targetting some of the measured statistics, doing whatever it takes to quickly improve such figures that are obviously correlated with the general welfare of the citizens. However, since they are politicians, the only means to their policies is resources forcefully taken from the public, and spent in unaccountable ways by a monopolist bureaucracy that no one chooses and that faces no competition. This unaccountability makes the delight of an army of parasites, amongst which the very economists appointed at various government-sponsored academic or political jobs themselves. And even when the money somehow escapes the parasites, unaccountability means there is no feedback loop between the effects and the decisions: how many parameters can you control when your only choice is to contribute a tiny fraction of the overall opinion in the marginal pick between two awful guys every so many years? And thus, all the money being spent is wasted for those who pay, in an anti-productive grab for the undeserved, as well described by the Law of Bitur Camember. The claim to improve the economy by taking resources away from said economy to fund wasteful programs that serve to breed parasitism is thus very much akin to drilling holes in order to increase the amount of cheese: it is a fraud that politicians convince themselves into believing so as to deceive the public more easily.
Of course, such intervention cannot succeed. By acting in expedient ways, politicians may for a short moment improve the statistics they target. But only for a while. For the very principle of their action is forceful destruction, and whatever improvement they may achieve by thus acting on a narrow measured figure can only be done at the detriment of all that is not measured, thus invalidating the very correlation that was implicitly or explicitly used to justify their political intervention: the inflated statistics that was targetted by the politicians thus becomes less and less meaningful as their intervention progresses. Finally, as their actions make the whole society gets poorer, politicians find that even those statistical figures that they desperately try to enhance start to decline.
Then, to hide their bad results even by the own crooked standards they chose, politicians will redefine their statistical measurements every so often, with new failures and redefinitions leading to further failures and redefinitions, progressively shifting the whole economical vocabulary into meaninglessness proportionately to the technicality of their statistics, all the while driving people to economic hardship. The end point of all this is a state where governments intervene a lot, spending a great amounts of resources for no positive results; what Anthony de Jasay calls The Drudge. Concerned people debate frantically to find a solution to their problems, but the words do not mean anything anymore because they have been stripped from their meaning, and thus all this frenzy leads nowhere either.
And it all ends when nothing is left but the rule of totalitarian politicians upon a stupefied mass of poor people. Whether this rule happens by the slow series of progressive reforms by stable parties, or whether some providential men are called to take the process to its ending more quickly, we can only wait to see what will happen in Europe first, and then in the USA. With a little bit of luck, the fate of the first countries that fall will scare the other countries out of taking the same path - but as long as people at large fanatically accept the statist ideology that crowns politicians as the only possible saviors who can solve all the problems of society if only they follow the right econometrical model as concocted by appropriate wise men, then people will ask ever more of the disease as the only medecine they can imagine.
Note: I first wrote about the these effects of statistics-based political intervention in section Statistics vs Cybernetics of my essay Government is the Rule of Black Magic. The illustration of the difference between correlation and causation based on the example of holes in the cheese was first told me by a good friend who seeks anonymity. I have since learnt that the phenomenon by which statistical measures lost their meaning when used by governments is already well-known and has a name: Goodhart's law (Cám ơn, Patri).
Finally, speaking of swiss cheese and government, here's the first result Google found me for "swiss cheese holes". No legislative gap will be left unfilled!