27 September 2010

Small Business Regulation Cost Study Off Base

The 2010 report of the Congressional Budget Office under the Regulatory Right To Know Act has been released and garnered an op-ed in the Wall Street Journal today by Nicole V. Crain and W. Mark Crain.

Crain and Crain Make Dubious Estimates Of Total Regulatory Costs

The WSJ op-ed inflates this number dramatically (to $1.75 trillion in costs per year), and argues that regulations unduly burden small businesses because their adjusted estimate of regulatory costs is higher per employee based on a report that they wrote for the U.S. Small Business Administration Office of Advocacy.

In contrast, the CBO report itself finds that "The estimated annual benefits of major Federal regulations reviewed by OMB from October 1, 1999, to September 30, 2009, for which agencies estimated and monetized both benefits and costs, are in the aggregate between $128 billion and $616 billion, while the estimated annual costs are in the aggregate between $43 billion and $55 billion."

The WSJ authors also make no effort to quantify the benefits associated with the regulations in question.

They argue that "The annual cost of federal regulations in the United States increased to more than $1.75 trillion in 2008, a 3% real increase over five years, to about 14% of U.S. national income," based on a regression models across countries that uses only a handful of inputs and a "regulatory quality index" that measures "perceptions of the ability of governments to formulate and implement sound policies and regulations that permit and promote private sector development" and an "unobserved components model." It comes up with a number wholely unconnected to actual directly measured costs of regulatory compliance, and doesn't even purport to be based on measures based on actual cost.

They also make it their policy in their report to use the highest possible estimate of every cost, despite acknowledging that this is controversial: "This report uses the high end of the cost range provided in the OMB reports and Hahn and Hird (1991)." Their estimate of regulatory cost for environmental regulations would be 19% lower if mid-point estimates were used instead, which is the only statistically defensible stance to take, because even if some costs are overestimated, the claims that costs are always underestimated doesn't hold water.

The cost-benefit analysis is inherently limited, because it depends upon how you choose to value lives saved and injuries avoided by regulatory efforts. But, the only way to conclude that the regulatory cost driving environmental regulations do not provide greater benefits than costs is to value lives lost about a sixth as much as the CBO does.

Measuring total costs and measuring benefits is inherently nebulous. There is a tendency to overlook or undervalue important benefits, and to count as the cost of actions that would have been required even in the absence of federal regulation by tort laws and other requirements as costs caused by the regulations. But, it can guide our considerations of public policy issues.

Dubious Allocations By Industry and Firm Size

The WSJ discussed study breaks regulatory costs into four categories "economic," "environmental," "tax" and "occupational health and safety" regulations.

Environmental, tax and occupational health and safety regulations are deemed more expensive per employee in small businesses in their report than large ones, although only the differences in cost per employee in environmental and tax regulations is particularly stark.

In reality, the only class of regulatory costs where the analysis is well supported by the evidence is their allocation of tax costs.

An examination of the expenses claimed as tax deductions by small businesses, for example, shows that one can allocate only a tiny percentage of those expenses to regulatory compliance costs. Many small businesses have fewer expensive overall than the amount of regulatory expenses that they are said by the WSJ authors' study to incur.

Cost of goods sold in a retail business, for example, simply doesn't constitute a regulatory cost under any reasonable view of the concept.

Small Businesses Reportedly Have Lower "Economic Regulation" Costs

Their study finds that compliance with "economic regulation" per employee, computed based on the "unobserved components model" grows as businesses have more employees and that these regulations make up more than half of all regulatory costs. So, they find the largest part of regulatory costs favors, rather than hurts, small businesses.

There is good reason to think that the entire concept of "economic regulation" as used in the WSJ authors' study is flawed. But, even if it is not, they do not make the argument that these regulations impose an unfair burden on small business relative to other businesses.

Small Businesses Have Higher Per Employee Tax Compliance Costs

The tax estimates, based on the actual forms filled about by firms by size and the actual time it takes to complete those forms and estimates of the cost of hiring tax preparers to complete them, are fairly solid. There is a simple economy of scale factor at work in this situation. Preparing a tax return for a business involves similar activities in big firms and small ones, and so economies of scale favor big firms on this count.

The economies of scale differences are also mostly gone once businesses have twenty or more employees. Most of the diseconomies of scale (about $750 per employee) involve only the smallest businesses.

Occupational Health and Safety Costs Differ Only Modestly By Firm Size and The Allocation By Industry In The Study Is Questionable

Occupational health and safety costs are allocated just $171 dollars per employee higher in the smallest businesses than in the average business.

Moreover, the allocation of occupational health and safety costs they make implausibly states that the cost of compliance is the same in every sector of the economy. In fact, it is almost certainly the case that compliance costs are much higher in industries like manufacturing, construction and utilities that make up a small share of all small businesses, than in trade, services and health care industries where small businesses are concentrated.

Occupational health and safety costs are roughly proportional to work related accidents that are uniquely work place related (as opposed for example, to work related crime and on the job traffic accidents in ordinary vehicles), and to OSHA inspections, not to pure employment, the basis used to allocate those costs by the WSJ authors.

One also gets a sense of how those costs differ from industry to industry by looking at worker's compensation insurance rates which are much higher in manfacturing, construction and utility industries than they are in commercial or health care environments. OSHA compliance in small businesses outside high risk industries in negligible.

Environmental Costs

The WSJ authors' study, concludes contrary to the CBO, that "economic regulation" of types unspecified, is a greater cost than "environmental regulation. The regulatory biggest benefits and costs according to the CBO report itself were: "the air pollution rules from the Environmental Protection Agency (EPA) produced 60 to 87 percent of the benefits and 58 to 64 percent of the costs."

Those rules predominantly impact utilities, car makers, and smoke stack manufacturing industries (a.k.a. heavy industry). Exceedingly few small businesses must comply with the Clean Air Act, and its benefits are well demonstrated in scientifically rigorous studies.

The main small business sector that produces air pollution and water pollution, farming, benefits from broad exemptions to the Clean Air Act and Clean Water Act.

Environmental regulation costs vary dramatically by business sector. The WSJ authors study breaks the economy into five business sectors: Manufacturing (5.5% of employment in firms with fewer than 20 employees), trade (with 19% of that employment), services (with 48% of that employment), health care (with 12% of that employment), and "other" with 16% of that employment.

They then use regression models for the business sector as a whole to allocate costs by firm sizes. The vast majority of environmental regulation costs are in the "manufacturing" business sector and the "other" business sector that includes construction and utilities.

There are no environmental regulation costs in any firm in the trade sector under their model, there are only $25 of environmental regulation costs in the service sector per employee under their model in firms with fewer than 20 employees and less than $203 of envirnomental costs per employee under the model in the health care industry in firms with fewer than 20 employees, sectors that combined make up 79% of all employment in firms with fewer than twenty people.

So, the vast majority of small businesses have negligible environmental regulatory costs.

Environmental costs per employee are greater in small employment manufacturing operations than in large ones, but the economic scale of manufacturing firms isn't particularly closely tied to the number of employees it has, due to automation, and because the environmental costs per employee vary dramatically by industry a manufacturing industry wide regression formula isn't particularly helpful in estimating environmental costs in small manufacturing firms which are outliers in many respects in a big business dominated industry.

A more relevant measure in the manufacturing industry, not examined in the WSJ authors' study, is environmental regulatory costs relative to gross revenues from sales. If environmental regulatory costs at a factory with fifteen employees (perhaps a microbrewery), are similar as a percentage of sales to those of a large factory in the same industry (perhaps Molsen-Coors), we shouldn't be concerned that there is any bias against small manufacturing businesses.

It also isn't a safe assumption that regression models showing environmenmtal costs per employee in medium and large firms hold true for smaller firms which may be in different industries, and for which the "fixed cost" factor of compliance in a regression model will tend to be inaccurate because different kinds of manufacturing equipment is used in small firms. The types of manufacturing industries most affected by environmental regulations tend to be large employers in heavy industries, and they tend to incur most of their costs generating air and water pollution, while small, light manufacturing operations tend to use electricity and natural gas purchased from utilities to carry out their operations.

The "other" business sector lumps together the utility and construction industries and then uses a similar regression model to that of the manufacturing industry to allocate costs, which is particularly inappropriate in that diverse business sector.

The methodology used allocates costs from high environmental regulatory cost utilities to low environmental regulatory cost small firms, mostly in the construction industry, distorting the figures dramatically.

The study's conclusion that firms of fewer than 20 employees in the "other sector" comprised largely of small construction firms, are incurring $13,760 per employee of environmental costs, which powerfully drives the purported environmental costs of small firms generally, is absurd.

The vast majority of environmental regulation costs derive from Clean Air Act and Clean Water Act regulations. Small construction firms generally do not directly comply with either. Costs incurred by tool making companies to make their tools and equipment greener are accounted for in the manufacturing firm side of the ledger, not at the construction firm level.

Also, small employment utility firms tend to be either rural electric cooperatives and renewable energy power producers.

Rural electric cooperatives generally incur little or no environmental regulatory costs themselves, because they usually purchase electricity from large centralized energy producers where the regulatory costs are appropriately counted because the employment activity and expenditures associated with the regulatory costs take place there. To the extent that environmental regulatory costs associated with electricty production are allocated to rural electric cooperatives, this should be measured relative to the cost of the electricty purchased, not relative to rural electric cooperative employment.

Small renewable energy based operations such as solar or wind power production firms also incur little or not regulatory costs under the Clean Air Act or Clean Water Act. Environmental costs incurred, for example, in manufacturing solar panels, are counted at the usually large firm manufacturing level in any case.

Of course, even if manufacturing companies and utilities with under twenty employees do have very high environmental regulation compliance costs per employee, that is a conclusion with a very different policy implication than a general statement that environmental regulation compliance costs are a drag on the small business sector generally. It has nothing to do with the regulatory environment of the vast majority of small businesses. It has a great deal to do with the sensible way to cope with the reality that technology has dealt us.

To the extent that environmental regulatory costs are higher in small utilities than in large utilities, it isn't obvious that the public should care. There is no great moral virtue to having small businesses in the utility industry, if power that produces similar environmental impacts can be produced more cheaply in larger firms. Utility regulation exists precisely because the technologically driven economies of scale associated with large operations, even to the point of having a near monopoly producer, have long been recognized.

Creating exemptions from environmental regulations for small firms in the same industry can create perverse incentives that make the regulation of large industries far less beneficial by shifting production to less efficient firms that generate more pollution.

Implications Of A More Accurate Analysis

More sensible allocations of regulatory costs by firm produce very different conclusions on what kind of agenda for regulatory reform small business would benefit the most from putting into place. Environmental regulations and workplace safety regulations start looking far less consequential to small businesses in general and even to small businesses in business sectors where those regulations are relevant to some extent.

"Economic regulations," whatever those are, don't appear to be particularly burdensome for small businesses.

Under a more realistic analysis, the main conclusion that applies across the board to small business regulation is that tax simplification confined largely to the smallest businesses is the place where there is the most opportunity to help small businesses by easing regulatory burdens, this is where the compliance costs are highest relative to the amount of tax collection involved.

There are ways to achieve this and a number of them have been recommended to the President by a panel of advisors. Simplifying tax law in areas like the home office deduction, deductions for the business use of cell phones, and reducing the distinctions between 1040 tax return income and self-emploment tax income can make a big difference on this front, as could simplified withholding tax rules for small employers. There is probably merit in simply removing a small amount of self-employment and tip income from the tax base entirely because it takes more work to collect than it is worth in tax revenues.

This is a very different conclusion than the one pushed by Crain and Crain who are effectively apply to make a "strave the beast" kind of anti-regulation argument to regulation in general, rather than examining in a fine tuned ways that regulations that may actually be a problem.

Marginal New Regulations

None of this means that it isn't possible that some regulations produce few benefits and high costs relative to those benefits.

Regulation is not inherently bad or good. A regulation that effectly puts an industry out of business can be a good one if that business produced a lot of harms and few benefits to society. There is no magic formula to determine how much regulation there should be in society any more than there is a formula to determine from first principles how many cars the automobile industry ought to sell each month. The more regulations that there are that provide more benefit than cost, the better. The fewer regulations that there are that involve more costs and benefits the better. The whole point of cost-benefit analysis is to provide an analytical framework to determine whether particular regulations should or should not be adopted, and to encourage regulators to fine tune regulations that create costs the the harms that they prevent.

Cost-benefit regulation is analysis isn't designed to quantify aggregate regulatory costs in an economy, something that is inextricably intertwined with the entire method of producing value in a healthy economy. For example, accounting system costs help managers make good business decisions, help banks make good financing decisions and help tax authorities collect the proper tax amounts and can't be entirely allocated to any one purpose. Some of the costs of tax compliance as a whole include costs that would be incurred by well run businesses even in a tax free world.

Even when benefits and costs aren't quantified in dollar terms, it makes sense to look at both in the case of individual regulations, recognizing that some costs and some benefits may be hard to monetize.

The CBO study identifies four individuual safety regulations that deserve examination:

The rules in the last year judged the most marginal by the CBO study involved (1) regulation of salmonella in eggs, (2) rules giving heightened strutiny to new trucking company compliance with safety rules, and (3) rules regarding stopping distance requirements for semis, each of which was deemed to have negative net benefit.

In addition, roof crush resistance for vehicles, which was determined to cost about $6.4-$11 million per life saved, which is higher than the value of lives saved usually used to estimate the benefit of new regulations.

It is a lot less headline grabbing to say that a handful of obscure rules promoting food safety and traffic safety, which everyone agrees are good things in the abstract, are really cost effective ways to meet their larger goals, than to rant about the specter of the growing regulatory state. But, if we want to be serious about having the right amount of regulation, rather than simply disliking regulation in all circumstances, even when it provides great benefits at low costs, this is precisely the kind of debate that we should be having.

1 comment:

Unknown said...
This comment has been removed by a blog administrator.