My first attack in the You Libertarians are so Impractical! series is on Sathish VM. Last heard, he was arguing with me that one cannot decide whether killing a lot of people is wrong or not unless we know whether the decision to kill the people was democratically taken or not. Now in his blog, he seems to be tired of seeing that though more and more rules are being made, companies are finding ways to wriggle out of them. So he has proposed a Pareto test rule. In his words
We can use the pareto test here? Only those corporate activities that bring the maximum wellbeing to all stakeholders, without making even any one of them worse off – is legal. Anything else is unlawful, not just immoral! Will this work?
Hm… let’s see. Taking a decision to revise salaries is a “corporate activity”. Is a decision to increase salaries pareto optimal? Of course not. Increasing salaries would make the shareholders worse off while making employees better off. So increasing salaries is out.
But wait. Increasing salaries is Pareto optimal in one case. If the employee is on the verge of quitting, and if it can be proved that the resignation would cause a significant loss in profits then the decision to increase salaries would be pareto optimal. ( I am ignoring the other stakeholders. Including them will only make the calculus more complicated and will probably mean that no decision can be taken at all)
But in which case can you prove that an employee is so valuable that his quitting will lead to a loss of revenue? A worker low-down in the scale can easily be replaced. But a manager higher up is valuable. So Sathish VM’s rule essentially ends up screwing the happiness of the ordinary worker at the expense of the shareholders and managers.
Sathish’s proposal is in fact so impractical that he must be a libertarian and an evil capitalist at heart.
Ravi, though I admire you sentiment, I think you are making the same mistake as most interventionists do – you are conducting a static analysis of the event.
An increase in wages maybe beneficial to shareholders, it might provoke a better work ethic or attract more efficient employees, possibly improving shareholder value.
I’d say a Pareto rule would not work, because Pareto was working and thinking in a highly idealised mode, where most of the world stands still while a certain act takes place. No single action taken by any player – pvt. or govt.- can be completly pareto optimal. The entire market process though leaves society better off as a matter of course.
I think layoff decisions are an easier example for corporate activities that have a benefits imbalance even though they might be the “right” decisions.
But Ravi has this habit of shocking you with a premise and holding on to it for pages and pages! 🙂
I see your point Gautam, but the point is, when you make something a legal requirement (on the pain of being sued and sent to jail) you will take decisions only on hard data. How are you going to prove “an improvement in work ethic” in a court of law? May be you can if you cook up the right data, but if you are going to be sent to jail or slapped with a fine if you fail to prove it, are you going to take a risk doing complicated dynamic analysis?
So that’s actually three strikes against Sathish’s proposal:
1) Unclear and overarching laws -> chilling effect on all decision making
2) Ex post facto laws -> you can’t figure out whether you did the right thing or not till you see the results of what you did.
3) Unintended consequences -> Already explained.
Ravi, can you take this up next?
http://www.savingcapitalism.com/capintro.pdf
I think Raghuram Rajan and Luigi Zingales have raised valid and important points regarding the need for government intervention.
In fact, just after I read this, I met some people from Asahi who said, automtive glasses are safer today thanks only to government regulations. Left to market, it might have taken more time.
(by the way, ‘a href=’ doesnt seem to be working)
Hi Ravi,
Thanks for discussing the Pareto Test idea. My question – “Will this work?” now seems to have an answer. 🙂
Regs,
Sathish