This is an anecdote.
The other day I went to the local town hall to get my passport renewed. In order to do so you have to bring “professional” identity photographs of yourself that they examine very carefully. Many people “fail” the test and are sent back to make new photographs. When I asked the lady who was handling my application why they do not take photographs themselves with a digital camera, she pointed to an empty metallic piece which once supported a digital camera, and next to it there was a note saying that since a government decree dated nov 2011, the administration will no longer take photographs itself. Instead people have to use the services of a professional photograph, although there are also automatic machines that one can use.
The government had bought thousands of high resolution digital cameras to produce the identity pictures of the millions of cititzens who ask for driving licenses, passports and identity cards. Shortly after, the professional photographers lobby complained that this was destroying their business. The government immediately yielded and issued a decree withdrawing the cameras it had bought and sending people back to the photographers.
Assume some fifty million french people have an identity card and another 25 million have a passport. These are 75 million documents that have to be renewed every ten years, hence 7.5 million documents being produced by the government each year.
Instead of the official losing five seconds to take the photograph, people have to go to an automatic machine or a professional photographer. Say they lose half an hour, at an average age of 15 Euros per hour that would be say 7 Euros lost. To this we have to add the cost of making the photograph, which is 5 Euros in a machine and 10 Euros for a professional photographer.
As an aside, the machines price discriminate: an identity photograph costs 5 euros, while other types of photographs only cost two euros. A textbook example: the first kind of customers is stuck while the other customers could use their digital camera, mobile phone, or just give up making a photograph instead.
Going back to our computation, let’s say 1/3 of the people use a professional photographer and 2/3 use a machine.
We end up with a financial cost of 1/3*7.5*10+2/3*7.5*5 = 25+25=50 million euros, to which we have to add the time cost of 7*7.5 = 52.5 million euros.
The annual welfare loss of this small, unnoticed, decree is therefore some 100 million euros. This has to be multiplied by the countless discretionary distortions that the government puts in place while nobody watches.
The arithmetics are simple: the cost of such a decree is diluted in the population, who loses some 2 euros per person. The benefits (a transfer of 1/3*7.5*10 = 25 million euros per year) are split between the relatively small group of professional photographers (to this we should add another 25 million to the shareholders of the firms that operate the machines, but we expect ownership of those firms to be diluted too). If there are say 10,000 of them, they each gain 2,500 Euros per year out of this lobbying success.
Unlike a law, a decree is not subject to parliamentary scrutiny. If what is going on is that the incumbent government would have lost the votes of the photographers had it not yielded to them, and does not lose any votes by protecting their rents because the losses per capita are small, then we may speculate that the parliamentary majority would have acted in the same way as the government.
On the other hand, the parliament is more likely to see the broader picture, and above all it passes laws, meaning it is capable of committing future policy. This commitment power would allow it to ban such discretionary transfers of rents to special interests, because it would recognize that when pervasive, they make everybody substantially worse-off. By contrast, the government, when contemplating to issue a decree, only sees one issue at a time and is more tempted to yield to the special interests. Thus the discretionary character of governance by decree increases the incentives of organized interests to engage in rent-seeking activities, with the usual adverse consequences for GDP and welfare.