Nation of Beancounters

The economics of safety regulation (pedagogical)

Posted in Explained from Scratch by Navin Kumar on June 7, 2012

Imagine a factory with a $10,000 monthly payroll: 100 workers making $100 each. There are several other factories further down the road, each paying the same wage. There is a (small) chance that a fire might break out, killing everyone. Workers are identical and are willing to pay 5% of their income – $5 each – to reduce the risk of death due to fire to zero.

Now imagine that a ‘fire department’ can be set up in-factory for a mere $250 per month that does just that. Is there a need for the government to pass a law ordering that all factories to set one up? No: the cost works out to $2.5 per worker – significantly less than what they’re willing to pay. If the workers are honest, all you have to do is – once a month – pass around a hat with a note saying “put two-fifty here to protect against fire” and hand the collection over to firemen. Everybody wins.

But wait! Once the fire service has been set up, it is impossible to exclude workers who don’t put money into the hat from enjoying the benefits of having resident firemen. By slyly passing along the hat without dropping any money in, I get the service for free. But if everyone thinks like me, no-one pays, the hat comes back empty and there’s no fire service. This problem of free riding is a major concern in public goods problems. In a small community, this wouldn’t be a problem: social pressure could be used to force people to contribute. But in a factory with thousands of workers, co-ordination becomes impossible.

Should the government make it mandatory? Not necessarily. Imagine that you’re the owner of the factory. If you don’t give the workers (atleast) $100 in compensation monthly, they’ll quit and  take a job at the other units down the street. But you don’t have to pay them entirely in cash: if you reduce their pay to $95 per month but start a fire service, the compensation stays at $100 – $95 cash and $5 of safety. Meanwhile your costs fall – you now spend $9750: $9500 in wages and $250 on firemen. Your monthly profits are $250 higher. It’s a win-win.

(In the long run, of course, competition between factories will mean that wages rise to $97.5, wiping out extra profits and returning the payroll costs to 10K per month: a situation indistinguishable from the one where a hat gets passed around and everyone honestly contributes.)

The general lesson here is governments don’t have to force producers to adopt safety measures. If workers value those safety measures more than they cost and markets are competitive, the problem will solve itself. 

Now, imagine that the service cost $750 per month instead – the workers simply aren’t willing to pay the required amount ($7.5 each) at their level of income. Fortunately, they’re in an industry where wages are rising. When their monthly wage rises to $150, it becomes worthwhile to start a service. The second lesson here is as incomes rise, so will safety standards.

But that’s horrible, say tender-hearted people. What if their incomes don’t rise – should they run the risk of being burnt alive just because they’re poor? Let’s pass a law ordering factories to start an in-house fire department. It’ll improve the welfare of the workers.

Actually, it won’t. Trade-offs exist. You could minimize the risk of dying in a car crash by never getting into a car. You could minimize the risk of drowning in the ocean by never going to the beach. But the benefits of engaging in these tasks exceed the potential costs. Similarly, you could buy insurance against exotic diseases – but you’d rather spend the money on a book or bread or a pen drive. When we say that the worker values the fire department at “no more than” $5 per month, we mean that the benefit he gains from spending $5 on something else – rice, medicine, his children’s education – exceeds the benefit he gains from the existence of the fire department. If we force the workers to spend $7.5 per month on a service that’s worth only $5 to them, they are made worse off.

But hold on, say the labour activists, we won’t make the workers pay for the fire service. We’ll make the factory owner pay for it. It’ll reduce his profits, but so what? The workers will be made better off.

Actually, no. Despite what the politicians say, you can’t impose a cost “on the employer” or “on the worker”. Some of the cost of the fire department will be borne by the employer in the form of lower profits and some will be borne by the employees in the form of lower wages and, in the long run, there is simply no way to legislate how much will be borne by who. Let’s look at both extreme cases.

If the cost falls entirely on the worker, he is made worse off for the reason mentioned above.

But let’s say that the costs fall entirely on the employer. Should you then force him to start a fire department? The answer is still no. Imagine that instead of ordering him to do so, the government were to tax him $750 and then distribute the amount among the workers. Their welfare would improve in both scenarios – but more so in the case where they get straight cash transfers since they value the $7.5 cash they receive more than the fire service they would’ve received, which is only worth $5. So even if you wanted to improve their welfare by transferring wealth (whether from customers or taxpayers or employers), there’s no reason to assume that the best way to do it is through safety regulations: cash transfers might be a better option. The third lesson is if workers don’t value the safety measure, mandating it would make them worse off.

The other half of the argument

Does this mean that safety regulations are never justified? Absolutely not! Till now we’ve assumed markets work smoothly, which in the real world, they rarely do. There are, broadly, five well-documented classes of market failure: market power (monopolies, oligopolies etc), externalities (pollution etc), imperfect information, self-fulfilling prophesies (bubbles, bank runs etc) and macro-economic fluctuation (recessions, depressions). The last two, while intimately related to each other, are not relevant to this specific problem. There are also many other sources of ‘failure’: the inability to make credible commitment, myopia etc but these are less well-documented, so we’ll be avoiding them today. Let’s discuss the first three.

Market power: A buyer that has market power is called a monopsonist – not to be confused with monopolist, which is a seller with market power. In the labour market, monopsonists reduce wages to below market levels in order to boost profits, at the cost of workers (just like monopolists increase prices to boost profits at the cost of customers). This also reduces employment as workers are less willing to work at this wage. In such a scenario, mandating safety measures increases the compensation that workers receive and cause both employment and output both rise.

This comes with a few caveats though – first, you have to prove that a monopsony exists. Let’s assume that’s done. You could also increase compensation by imposing a minimum wage on the industry: raising the “cash” component of their salary. Whether this option is superior to imposing safety regulations or not depends on what the workers want, which can be difficult to figure out. A better option might simply be to force unionization, giving the workers leverage over their employers and raises wages and employment (at least in a monopsonistic market). Since strong unions can collectively bargain for their members and force them to pay dues, this may solve the empty hat problem as well. (It should be kept in mind that the overall impact of unions on public welfare is controversial.)

Imperfect information: The factory owner claims that those men wearing fancy red coats are firemen – but how do you (the worker) know they’re not unemployed actors who would be completely useless in a real fire? You don’t have the expertise to judge. Meanwhile the factory owner is offering only $95 while the factory down the street is offering $100, although there’s no promise of protection against fire. You decide that you don’t trust this factory owner and take a walk. The factory owner, realising that no-one wants to work for him at this particular compensation package discontinues the department (which turned out to be real after all) and goes back to paying employees $100.

But let’s say that workers do think that the factory owners are being honest and work for a $95 wage – this gives crooked factory owners an incentive to hire actors, boosting their bottomline. After a fire or two, the workers stop believing in the innate goodness of mankind and we wind up back in the preceding paragraph, where even honest factory owners cannot implement fire departments. There is no equilibrium in which factory owners will honestly acquire a fire department and workers will take the required pay cut.

If the workers cannot verify that the safety measures have been implemented, sub-optimal outcomes are possible. In such a scenario, it could be a win-win if the government starts enforcing safety measuring with standards and inspections: workers are assured of a safe working environment and are therefore more willing to take lower wages. And employers can now offer lower wages since they have the government making sure they keep their word about those firemen.

Unfortunately, there are a number of plausible objections to this. Firstly, it requires workers to trust the government. Inspectors can be bribed and maybe the government isn’t really interested in working conditions – merely in scoring political brownie points.

Secondly, a private agency – which has a reputation on the line – can also do the inspection and certify to workers that the employer is good for his word. There’s no need for the government to do the job. Alternatively, safety measures could be coded into a contract so that if fraud is discovered, the employees can take the employer to court. (This argument is problematic in a country where trust of business is low and courts are inefficient).

And, finally, this is NOT an argument for the government creating and enforcing standards – it’s an argument for the government policing existing standards – making sure that the employers are doing what they claim to be doing (though one can make the argument that it’s easier to enforce a standardised set of rules).

Imperfect information, unlike market power, isn’t a very good argument for heavy government intervention (except as a “second best” solution, which no one seems to be claiming). At best, it’s an argument for improved policing.

Externalities: Imagine that, instead of a single factory in an area, you had an industrial cluster with hundreds of units. A fire could easily spread from one unit to another. Extinguishing a fire in one unit benefits not only that unit, but also the one next door. Imagine that there are 100 units (identical to the one at the beginning) and a fire department that can serve the entire place at a cost of $25,000. No single unit would bother setting up such a fire service*, but it would be worthwhile for the cluster as a whole to do so. Unfortunately, the co-ordination and free riding problems raise their ugly heads again – and this time there’s no common payroll that can be used to adjust everyone’s cost. In a scenario like this, the government can potentially improve matters by taxing all units $250 and spending it on a common fire department, greatly increasing public welfare.

To sum up: there is no need for government intervention to improve safety standards if markets are competitive, and mandating a safety measure can be harmful (or atleast less beneficial than a cash transfer) if workers don’t care for the measure. As incomes rise, standards will improve. Finally, these lessons are not true in all cases: if markets aren’t competitive, if information isn’t perfect or if externalities exist, a case can be made for intervention.

*if a fire is raging and spreading from a neighbouring unit, an “in-factory” unit would be too small to do anything.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: