The increase in the number of federal register pages is a response to court cases in the 1970's that made it easier to challenge federal administrative agency actions on the basis that they were inadequately justified. The 1935 Administrative Procedure Act requires agencies to accompany regulations with a "concise general statement" in the Federal Register. From 1935 to the 1970's, that's precisely what they were, even when the substantive regulations had far-reaching scope.

In the 1970's there was a trend of increased judicial scrutiny of agency actions, centering on the agency's explanations and rationales for new regulations. Sometime in the 1960's, economic cost-benefit analysis also became a fixture of agency action. Thus, the page count of the federal register exploded, as the "concise general statement" became exhaustive rationales filled with the results of studies, cost-balancing analysis, etc.

The end result is that agencies like the NLRB have vastly less actual power than they did decades ago, but to exercise what little power they have, they must detail every bit of reasoning and evidence behind a rule in the Federal Register, to fortify the new rule against challenges in court. Modern agencies have to slavishly justify every little action, whereas agencies of the 1930's-1960's could impose sweeping regulations of industries without so much as a cost-benefit study.

The "greatly increased compliance costs" narrative is garbage. One would assume that if compliance costs were increasing, legal costs would increase too. Yet, the opposite is true: the legal sector's share of real GDP has shrunk from over 2% in the 1970's to 1.3% today: http://amlawdaily.typepad.com/.a/6a00e55044cbaf88340168e5077.... Indeed, that understates the decrease. Finance is a very legal-intensive industry, and makes up a much larger fraction of the economy than it did in the 1970's. Thus, the decrease in legal expenditures in the non-finance sectors must have been even larger.

Now, legal expenditures are not necessarily the same as regulatory compliance costs, but one would imagine the two are highly correlated. And the trend is totally in the opposite direction from the "greatly increased compliance costs" narrative.

If you are right, the Small Business Administration (a government agency itself,) must be very wrong, as it found "[t]he annual cost of federal regulations in the United States increased to more than $1.75 trillion in 2008." [1] The United States' GDP was $14.22 trillion in 2008, meaning that 12.3% of GDP was being used to comply with federal regulations, if you were to add state regulations, the result would only become more staggering.

[1] http://www.sba.gov/sites/default/files/The%20Impact%20of%20R...

$1.3 trillion of the number comes from estimating the difference between theoretical GDP and actual GDP given "economic regulations," ranging from quotas and tariffs to the Americans with Disabilities Act. They use a regression analysis based on a "Regulatory Quality Index" published by the World Bank. I don't know what to call this methodology other than "garbage." Fundamentally, many of these things are legitimate trade-offs, and it makes no sense to conflate those trade-offs with "compliance costs." A world in which the disabled are excluded from society is almost certainly going to be one with a higher GDP. A world in which we import all our food from South America is almost certainly going to be one with higher GDP than one in which we subsidize farming in the U.S. The fact that we choose not to live in that world doesn't make the difference between that theoretical GDP and the actual GDP a "compliance cost." Now, the cost of ADA litigation, or the cost of agricultural subsidies, etc, might fairly be considered a "compliance cost" but the SBA analysis goes far beyond that.

It also analyzes things like the increased cost of electricity resulting from environmental regulation, without accounting for the cost of environmental damage. A Harvard study found that coal use alone causes $350 to 500 billion in externalized health costs: http://www.fastcompany.com/1727949/coal-costs-us-500-billion.... Under the methodology of the SBA study, a $100 billion increase in electricity costs from banning coal would be counted as a "compliance cost" but the $300-500 billion saving from reduced health damage wouldn't be accounted for.

This is a very important methodological flaw, because most environmental regulations are fundamentally about dealing with negative externalities, by forcing them to be internalized. The SBA's methodology will account for the internalized costs, but not for the net economic benefit that arises from eliminating the externality. It's fundamentally flawed from a purely economic theory standpoint.

NB: I'm actually supportive of the general trend of deregulation, although I think gutting environmental and antitrust enforcement is a bad idea. But the trend is indeed one of deregulation, not increasing regulation.