How Do Federal Regulations

Affect Consumer Prices?

An Analysis of the Regressive Effects of Regulation

Dustin Chambers and

Courtney A. Collins

February 2016

MERCATUS WORKING PAPER

Dustin Chambers and Courtney A. Collins. “How Do Federal Regulations Affect Consumer
Prices? An Analysis of the Regressive Effects of Regulation.” Mercatus Working Paper,
Mercatus Center at George Mason University, Arlington, VA, February 2016.

Abstract

While several scholarly papers have documented potential costs associated with the burden of
federal regulations, none have provided a comprehensive empirical analysis of the effect of
regulations on consumer prices. This study examines the relationship between regulatory
expansion and higher prices and asks whether those price increases have a disproportionately
negative effect on low-income households. By combining microdata from the Consumer
Expenditure Survey with industry-specific regulation information from RegData and price
changes in the Consumer Price Index, we find evidence of regressive regulatory effects. Our
results suggest that the poorest households spend a larger proportion of their income on goods
that are heavily regulated and subject to both high and volatile prices.

JEL codes: I18, D12, H23, L51

Keywords: regulation, federal regulations, regressive effects, distributional effects, consumer

prices, Consumer Expenditure Survey, RegData

Author Affiliation and Contact Information

Dustin Chambers, PhD

Associate Professor of Economics
Salisbury University
[email protected]

Courtney A. Collins, PhD

Assistant Professor of Economics
Rhodes College
[email protected]

All studies in the Mercatus Working Paper series have followed a rigorous process of academic evaluation,
including (except where otherwise noted) at least one double-blind peer review. Working Papers present an
author’s provisional findings, which, upon further consideration and revision, are likely to be republished in an
academic journal. The opinions expressed in Mercatus Working Papers are the authors’ and do not represent
official positions of the Mercatus Center or George Mason University.
 How Do Federal Regulations Affect Consumer Prices?

An Analysis of the Regressive Effects of Regulation

Dustin Chambers and Courtney A. Collins

1. Introduction

The 2012 Code of Federal Regulations includes more than a million individual restrictions,

representing a regulatory burden that has grown by more than 28 percent over the previous 15

years (Al-Ubaydli and McLaughlin 2015). Certain industries have experienced even higher

regulatory growth over the same time period. For example, federal regulations related to

highway and street construction increased by 94 percent over the past decade and a half. The

natural gas distribution industry experienced a 109 percent rise in regulations, and the

corresponding increase in the water and sewage industry was 125 percent.1

There is substantial variation in the types of regulations that exist both across and within

industries, as well as in their numerous potential effects on consumers. This study focuses

specifically on how regulation growth affects consumers through its impact on prices. While

most regulations are not passed with the explicit goal of raising prices (and, in fact, some are

created specifically to decrease prices), compliance with regulations often translates into higher

costs for businesses, which in turn may drive up prices for consumers. If this rise in prices

occurs, regulatory growth is unlikely to affect all consumers equally. Because high- and low-

income families have different spending patterns, regulations that increase prices in a particular

market sector often have a disparate socioeconomic impact.

1
All estimates of the regulatory burden are from the RegData database of the Mercatus Center at George Mason
University (North American Industry Classification System [NAICS] 2212—natural gas, NAICS 2213—water
sewage, and NAICS 2373—highway and street construction).

3
 Recent information from the Consumer Expenditure Survey (CE) reveals that households

just below the poverty line spend a substantially larger percentage of their income on

transportation and gasoline, utilities, food, and health care than do high-income households

(Goldstein and Vo 2012). To the extent that, on balance, regulations raise prices, regulations will

cause regressive effects if they are concentrated in the economic sectors where low-income

households spend the most. The purpose of this study is to analyze the potential regressive

effects of federal regulations—first by documenting differences in consumer spending patterns

across income levels and then by examining how regulatory growth has affected the prices of

goods and services for consumers across the income distribution spectrum.

By using detailed microdata from the CE, we first assess whether there are meaningful

differences in the spending habits of average consumers from different income groups. We join

these data with information on regulatory restrictions by industry, available from the RegData

database of the Mercatus Center at George Mason University, and data from the Consumer Price

Index (CPI) to determine the effect of regulatory expansion on price levels. We allow for

differences in the inflation rate by consumer income group to examine potential regressive

regulation effects. We find evidence of a statistically significant relationship between regulation

and price levels: specifically, a 10 percent increase in total regulations leads to a 0.687 percent

increase in consumer prices. We also find that households from the poorest income groups

experience both the highest overall levels of inflation and the highest levels of price volatility.

2. Background on the Costs of Federal Regulations

Measuring the full costs of federal regulations is difficult. The Regulatory Right-to-Know Act of

1999 requires the Office of Management and Budget (OMB) to publish an annual report

detailing the costs and benefits of major federal regulations. In its May 2014 report, OMB

4
estimates the annual cost of regulations to be between $74 billion and $110 billion.2 However,

OMB openly acknowledges that this estimate is far from a complete approximation of all federal

regulatory costs. For example, the report excludes costs associated with rules that are more than

10 years old and rules that are not defined as major (i.e., rules that have an annual economic

effect of less than $100 million).

Crain and Crain (2014) estimate that the true comprehensive cost is more than $2

trillion,3 including all regulations and accounting for many indirect costs, such as reduced

economic productivity, that are absent from OMB’s analysis. The authors note that some portion

of these costs is passed on to consumers in the form of higher prices, although they neither model

nor empirically estimate this increased consumer inflation.

Several papers address the potentially harmful unintended consequences of regulations.

McLaughlin and Williams (2014) outline some of the adverse outcomes related to regulatory

accumulation, or the “buildup” of old or obsolete regulations inherent in the US regulatory

system, including lower rates of economic growth, reductions in the establishment of new

businesses, and reduced international competitiveness. There is a substantial literature illustrating

these regulatory consequences in both the United States and abroad.

For example, Dawson and Seater (2013) examine the specific impact of federal

regulations on economic growth and estimate that since 1949 increased regulation has

significantly decreased the rate of economic growth, resulting in an accumulated GDP loss of

$38.8 trillion by 2011. Other papers report a negative relationship between regulatory growth

and economic productivity, including Nicoletti and Scarpetta (2003); Djankov, McLiesh, and

2
This cost estimate is in 2014 dollars, as quoted by Crain and Crain (2014). The actual estimate cited in OMB
(2014) is $68.5 billion to $101.8 billion, in 2010 dollars.
3
This estimate is in 2014 dollars.

5
Ramalho (2006); and Crafts (2006). Gørgens, Paldam, and Würtz (2003) explore the possibility

of a nonlinear relationship between regulation and economic growth and find that the bulk of the

effect stems from a transition from moderate to heavy levels of regulation.

One key channel through which federal regulations are likely to affect economic growth

is by creating significant barriers to new business entry. Benson (2004) discusses this barrier as a

significant opportunity cost to regulation. Empirical studies find that increased regulatory start-

up costs lead to lower rates of new business entry in both Europe (Klapper, Laeven, and Rajan

2006) and the United States (Fisman and Sarria-Allende, 2004). Ciccone and Papaioannou

(2007) examine the time it takes new businesses to comply with regulatory entry requirements

and find that reducing red tape is associated with increases in the number of start-ups.4

A significant body of literature examines the potential unintended consequences and

costs of environmental regulations, specifically the Clean Air Act, the Clean Water Act, and their

succeeding amendments. Becker and Henderson (2000) show how the Clean Air Act altered

businesses’ decisions regarding the construction, location, and size of new plants. In response to

the new regulations, firms were more likely to build smaller plants in low-pollution areas.

Although the firms’ decisions were in compliance with environmental legislation, the costs of

building inefficiently sized plants in suboptimal locations were significant.5 Greenstone (2002)

documents substantial job losses, decreases in capital investments, and reduced output as a result

of the same regulations. Hazilla and Kopp (1990) emphasize the importance of accounting for

social costs when evaluating the effects of environmental regulations, rather than simply

including private expenditures. They highlight the potential spillover effects outside the industry

4
For other examples detailing the relationship between regulation and economic growth, see Ardagna and Lusardi
(2010) and Benson (2015).
5
For related research detailing the effects for specific industries, see Becker and Henderson (2001). Additionally,
Becker (2003) examines how local community attributes predict the level of investment in pollution abatement.

6
that are directly affected by the regulations and note that the social costs of regulation likely

increase across time.

An additional consequence of federal regulations is their potential regressive effects.

While there is a substantial body of literature on the regressive effects of taxation,6 few studies

explore the distributional consequences of regulation. Two exceptions are Crain and Crain

(2010) and Thomas (2012). Crain and Crain analyze the effects of regulations on businesses and

find that small firms bear a disproportionate burden of compliance costs.

Thomas (2012) argues that many health and safety regulations are regressive because

they target risks that often reflect the preferences of high-income households. Relative to their

low-income counterparts, high-income households have a stronger preference for reducing low-

probability risks that are costly to mitigate. When these risks are addressed by regulations, all

market participants (regardless of income) pay the cost—in the form of higher prices for

consumers and lower wages for workers. Thomas contends that regulatory costs are

disproportionately borne by low-income households, inasmuch as they are obliged to pay for

higher levels of health and safety than they would in the absence of regulation. In addition, these

costs potentially crowd out private risk-reduction spending by low-income households.

Miller (2012) allows for the possibility of distributional effects in her analysis of the

federal energy conservation regulation for new residential dishwashers. The Department of

Energy, which issued the new regulation in 2012,7 estimated that it would increase dishwasher

prices by 13 percent. Interestingly, Miller reports that the breakeven point for a consumer to

6
See Poterba (1991) for an analysis of gasoline taxes, Wier et al. (2005) for an analysis of carbon dioxide taxes, and
Borren and Sutton (1992) for an examination of cigarette taxes.
7
Department of Energy, Direct Final Rule: Energy Conservation Standards for Residential Dishwashers, RIN No.
1904-AC64, May 30, 2012, https://www.federalregister.gov/articles/2012/05/30/2012-12340/energy-conservation
-program-energy-conservation-standards-for-residential-dishwashers#h-12.

7
recover a higher dishwasher price from energy savings is 11.8 years of use, which is longer

than the average 9- to 12-year life span of a new residential dishwasher. Miller calculates that

the breakeven point for senior adults and low-income households is more than 13 years,

suggesting that these consumers are harmed even more than other households by the energy

savings regulation.

While studies such as Miller’s examine the effect of specific regulations on prices in

particular industries, no study to date offers a comprehensive analysis of the impact of

regulations on consumer prices in general. This paper contributes to the literature by empirically

estimating the relationship between increased regulations and inflation and by examining the

extent to which regulations are regressive. We begin by examining basic spending differences

across different income strata, using data from the CE and incorporating regulatory restrictions

from the RegData database. We then use the expenditure data to create basket weights to

construct several CPI-based price indexes. Finally, we use the price indexes in an analysis of the

effect of regulations on consumer prices.

3. Differences in Spending Patterns across Income Groups

Our fundamental argument is based on the assumption that low- and high-income households

differ in their spending habits. In particular, low-income households spend a larger percentage of

their income on particular goods and services relative to high-income households. Before

determining how regulations affect consumer prices and exploring any potential heterogeneity in

the effect for different types of consumers, it is important to document the differences in

spending patterns across income groups. Specifically, we are interested in whether spending by

low-income households is more heavily concentrated in consumption categories that are subject

to higher levels of regulation.

8
The Consumer Expenditure Survey

To answer this question, we combine two sets of data: public-use microdata from the CE and

industry-specific data on regulatory restrictions from the RegData database. The CE includes

quarterly interview surveys, conducted by the Bureau of Labor Statistics (BLS), of

approximately 7,000 US households. It is constructed as a rotating panel, in which each

household is interviewed once every three months for five quarters and then is dropped from the

survey. The survey contains information related to household income levels and demographic

characteristics, as well as detailed data that describe household expenditures.

The CE dataset is organized by the Universal Classification Codes (UCC) system, which

consists of six-digit codes that categorize goods and services into specific purchase groups.

Households are queried about the details of their monthly spending habits. Each purchase is

recorded and labeled with an appropriate UCC. The CE also includes income files for each

household. Matching the expenditure files with the income files allows us to examine UCC

expenditure habits by income level. We examine the spending activities of five income

quintiles—the lowest 20 percent of income earners, the second-to-lowest 20 percent of income

earners, the middle 20 percent of income earners, the second-highest 20 percent of income

earners, and the highest 20 percent of income earners. By aggregating monthly spending values

across the year and averaging by income quintile, we derive average annual spending by UCC

for each income quintile.

Merging these data with information on the regulatory burden for each expenditure

category allows us to determine if there are differences in spending habits in terms of regulations

between households of different income levels.

9
RegData

While regulations can be used to refer to the guidelines published in the Code of Federal

Regulations, it is important for our empirical work that we precisely define the term. Our

regulation measures come from RegData, the Mercatus Center’s database of industry-specific

federal regulations. RegData is unique in its method of measuring regulatory burden. It analyzes

rules and guidelines published in the Code of Federal Regulations, but instead of reporting page

counts or number of rules, it counts each specific binding restriction that appears in the text of

policies. Each time a policy includes a word indicating an obligation, such as must or shall, that

word is counted as a restriction.8 These restrictions are weighted by their industry relevance and

summed to produce a regulatory index value.9 Regulatory index values are reported by industry

and by year, so it is possible to track regulatory restrictions within a particular industry over

time. All our empirical calculations and estimates of “regulations” refer to this regulatory index

from RegData.

RegData reports regulations by two-, three-, and four-digit codes of the North American

Industry Classification System (NAICS). To combine this information with the expenditure and

income data from the CE, we link NAICS codes to UCCs using commodity input-output tables

from the Bureau of Economic Analysis and the Consumer Expenditure/Personal Consumption

Expenditure Concordance from the BLS.10

We have approximately 350 unique UCCs for each year. To create broader spending

categories—and to facilitate an eventual examination of the effect of regulation on prices—we

collapse UCCs into the basic CPI expenditure categories used by the BLS. Using the BLS’s CE-

8
Five words are coded as restrictions in RegData: shall, must, may not, prohibited, and required.
9
For details on the methodology of calculating measures of regulation, see www.regdata.org/methodology.
10
For a detailed description of the methodology mapping regulations from the NAICS space onto goods and
services in the UCC space, see appendix A.

10
to-UCC aggregation scheme, we match the 61 expenditure categories from the CE with our

regulation dataset indexed by UCC code.11 Our combined dataset includes data for the years

2000–2012.

By using income information available from the CE, we divide households into five

income quintiles. This division allows us to examine spending habits across a broad range of

income levels. Our measures of regulation include both direct regulations, which capture

restrictions affecting a good or service itself, and input regulations, which capture restrictions

affecting the supply chain of inputs for a particular good or service (see appendix A for details).

The variable total regulations is the sum of direct and input regulations.

Consumer Expenditure Patterns

Table 1 (page 24) shows the percentage of spending for each income quintile in categories with

very high and very low levels of regulation. These numbers represent average values for each

income quintile spanning the time period 2000–2012. Households in the highest-income quintile

make 54.5 percent of all their expenditures in the 25 most heavily regulated expenditure

categories, where regulations for goods and services are measured directly (excluding input

regulations). The corresponding number for the lowest-income households is 60.3 percent, which

is a 10 percent higher consumption share compared to high-income households. Including all

regulations, the relative difference is about 12 percent.

A mirror-opposite pattern is evident when comparing expenditures in the least regulated

expenditure categories. The highest-income group allocates 32.19 percent of its total spending to

11
As a starting point, we used the expenditure category to UCC mapping contained in the BLS’s Dstub2010.txt
aggregation processing file. For missing or sparsely covered expenditure categories, we used an additional
expenditure category to the UCC mappings. For more information on this file, see “2010 Consumer Expenditure
Diary Survey: Public Use Microdata, User’s Documentation,” http://www.bls.gov/cex/2010/csxdiary.pdf.

11
goods and services subject to the fewest direct regulations, while the bottom-income quintile

spends 25.64 percent of its total expenditures in the same category. Total regulations reflect the

same patterns, with high-income households spending more (38.6 percent) in lightly regulated

areas than low-income households (31.9 percent).

Table 2 (page 25) presents the expenditure categories for which the difference in

expenditure allocation between the bottom- and top-income quintiles is the greatest.12 These

are areas in which the lowest-income families allocate a larger share of their overall spending

than do higher-income families. These categories contain rent and utilities, including

electricity, telephone services, and audio and visual equipment and services. Households from

the lowest-income quintile spend more than five-and-a-half times as much on rented dwellings

than households from the highest-income quintile, as a percentage of total expenditures.13 They

spend almost 85 percent more on electricity as a percentage of total expenditures and 50

percent more on telephone service. Other areas where the poorest households spend a larger

proportion of their income are drugs and medical supplies, medical insurance, and

miscellaneous food items.

To explore the regulatory restrictions that apply to these categories, figure 1 (page 26)

plots annual direct regulations for each of these expenditure categories from 2000 to 2012.14 For

most categories, there is a general upward trend in regulations over the sample period.

Exceptions are the cigarette industry, which has experienced a downward trend (at least until

recently), and the category that includes medical services and insurance, which experienced a

12
For a complete list of the top 20 expenditure categories and their corresponding direct and total regulation ranks
for each of the five income quintiles, see appendix B.
13
Note that spending for each quintile is reported as a percentage of overall total expenditures for each income
group. The level of total spending in most categories is greatest for households in the top quintile.
14
RegData contains no federal direct restrictions for the nonalcoholic beverages expenditure category, so we include
no corresponding graph of changes in regulation for this category.

12
sharp initial drop in regulations, followed by a steep increase. The category containing rented

dwellings also experienced a recent spike in regulations, following earlier variation across time.

Most of the expenditure categories that capture basic utilities show substantial growth in

regulations: direct regulations for electricity, telephone service, and audio and visual equipment

and services all increased by 33 percent to 37 percent. Regulations in the gasoline industry grew

by 33 percent. The largest increase occurred in the drugs and medical supplies category, which

experienced an almost 90 percent increase in direct regulations.

Taken together, these data support the argument that there are important differences in

consumer spending patterns by income groups. We find that, relative to the wealthiest

households, the poorest households spend a larger percentage of their income on goods and

services that are more highly regulated and a smaller percentage of their income on goods that

are less regulated. There are particularly large differences in spending patterns for utilities,

including natural gas, electricity, and cable or satellite television service. The regulatory burden

for these industries has increased sharply over time. In most cases, these increases have outpaced

the overall average growth rate of all regulations.

4. Calculating Price Changes by Income Group

Given the established differences in spending habits across income groups, we seek to determine

whether increased regulations have a disproportionately negative effect on low-income

households in the form of higher prices for goods and services, which compose a large share of

their expenditures. To explore these potential regressive regulatory effects, we must link our

expenditure/regulation dataset with price changes across time. Because our data are organized by

CPI expenditure categories, we can easily merge annual CPI price levels into our existing

13
dataset. The BLS publishes expenditure data by household income quintiles for the same

categories, allowing us to examine differences across income groups.15

By using these data, we construct consumption expenditure basket weights for five

household income groups. We exclude nonconsumption expenditures (e.g., charitable

contributions, life insurance payments, and retirement contributions) and match the remaining 61

expenditure categories with the CPI price data.16 The resulting balanced panel contains price and

basket weight data for 61 categories spanning the years 2000–2012 (793 observations). Table 3

(page 27) contains the names of each expenditure category, the average basket weight by income

group, and the direct, input, and total regulations for each expenditure category. We also use the

expenditure basket weights to construct annual aggregate weighted regulation series for each

income group:
!
𝑅𝑒𝑔!! = ! 𝑤!" ∙ 𝑅𝑒𝑔!" , (1)

where 𝑤!"! are expenditure basket weights equal to the proportion of spending in year 𝑡 on

expenditure category 𝑖 by households in quintile ℎ (ℎ = 1, 2, … , 5), and 𝑅𝑒𝑔!" are the regulations

that apply to expenditure category 𝑖 in year 𝑡. Table 4 (page 29) reports the weighted regulations

that apply to the combined, all-households group.

By combining the basket weights and price data, we construct two alternative price

indexes for each income group. The first is a classic Laspeyres price index, whereby for each
!
income group (ℎ), fixed basket weights from the base year (2000), denoted by 𝑤!,!""" , are

multiplied by their corresponding current-year category prices (𝑃!" ) and summed over the

expenditure categories (indexed by 𝑖 ) to derive the following index:

15
See Expenditure Shares Tables by income quintile, 1989–2011: http://www.bls.gov/cex/csxshare.htm.
16
See Archived Consumer Price Index Detailed Report Information: http://www.bls.gov/cpi/cpi_dr.htm. For the
lowest-income quintile, the included categories covered 85.2 percent of expenditures in 2012. For the highest-
income quintile, said categories covered 79.9 percent of expenditures in 2012.

14
 !,!"#$%&'%# !
𝑃! = ! 𝑤!,!""" ∙ 𝑃!" . (2)

The widely used Laspeyres price index suffers from a number of well-known problems,

most notably substitution bias. To overcome this shortcoming, we calculate the following

chained price index:

!! !
!" !!!"!!
!!"
𝑃!!,!!!"#$% = 𝑃!!!
!,!!!"#$% !
× ! ! . (3)
!"!!

Table 5 reports the aggregate price indexes for both foregoing methodologies.

Interestingly, regardless of the index used, the rate of inflation is highest for the poorest

households, declining with increased income.

5. The Effect of Regulations on Prices

Price Levels

Figure 2 (page 30) provides a scatter plot of the weighted total regulations from each of the five

income groups against their corresponding group-specific chained price series. Clearly, there is a

strong positive correlation between total regulatory burden and total prices. That said, both prices

and regulations trended upward over the sample period (2000–2012), so it is important to explicitly

control for this common trend to rule out any spurious correlation. To do this, we compare the

growth rate of prices over time (i.e., inflation) against the growth rate of regulations.17 The simplest

model of this relationship is the following autoregressive time series equation:

! !
𝑝!! = 𝛼 + 𝛽𝑟𝑒𝑔!!! + 𝜌𝑝!!! + 𝑢!! , (4)
!
where 𝑝!! is the log first difference of the chained price series for household ℎ, 𝑟𝑒𝑔!!! is the log

first difference of the total regulations series for household ℎ lagged one year, and 𝑢!! is a mean

17
In practice, we transform the price and regulation data by taking their natural logarithm and first differencing each
series. This calculation effectively yields the growth rate of each series.

15
zero error term. Intuitively, this equation specifies that for a given income group the current rate
!
of inflation (𝑝!! ) is determined by the prior year’s inflation rate (𝑝!!! ), as well as the prior growth
!
rate in regulations (𝑟𝑒𝑔!!! ), which accounts for the natural lag that exists between creation and

publication of new regulations and their measurable impact on the market for goods and services.

The estimation results for equation (4) are provided in column 1 of table 6B (page 33). The

coefficient on lagged regulatory growth is positive and strongly statistically significant, equaling

0.0648, implying that a 1 percent increase in total regulatory restrictions increases consumer

prices by an additional 0.0648 percent.18

To ensure that these results are robust and that inclusion of a one-period lag (𝑡 − 1) of

prior regulatory growth is appropriate, we consider seven alternative specifications of

equation (4), which include every combination of the following three variables: current

regulatory growth (𝑡), a one-period lag (𝑡 − 1) of regulatory growth, and a two-period lag (𝑡 − 2)

of regulatory growth. The results are reported in table 6A (page 32). Without exception, current

regulatory growth and the two-period lag of regulatory growth are statistically insignificant in

every variant of equation (4) in which they appear. This result supports our earlier theory that

there is a natural gestation period between the publication of new regulatory restrictions and their

measurable impact on prices. After the impacted production processes have been altered to

comply with new regulatory dictates, there is an associated jump in the price of these goods and

services. Moving forward, these regulations do not promote additional inflation as their effect is

already captured in the change in the price level of the affected goods and services, suggesting

that longer lags of regulatory growth should not have a statistically significant effect on current

inflation. We also perform a lag selection exercise, examining the Akaike Information Criterion

18
We use White (period) robust standard errors throughout unless otherwise specified.

16
and the Schwarz Information Criterion for alternative versions of equation (4). The version of
!
equation (4) that includes only 𝑟𝑒𝑔!!! was selected by both the Akaike and the Schwarz criteria

as it possessed the lowest values for both.

One obvious shortcoming of equation (4) is that the rate of inflation for each income

group differs (see table 5). Therefore, the common intercept assumption of equation (4) should

be replaced with unique intercepts for each income group, as in the following:
! !
𝑝!! = 𝛼 ! + 𝛽𝑟𝑒𝑔!!! + 𝜌𝑝!!! + 𝑢!! . (5)

Equation (5) is a dynamic fixed-effect panel model. Unfortunately, standard fixed or

random effects methods yield biased coefficient estimates in such models. Therefore, we use

Arellano and Bond’s (1991) generalized method of moments (GMM) estimator, which was

specifically developed to estimate dynamic fixed-effect panel models. A brief sketch of this

estimation procedure will follow; those interested in a fuller exposition should see Arellano and

Bond. To begin, equation (5) is first-differenced to eliminate the income-group fixed effects.

Next, a suitable instrument set is constructed, consisting of lagged predetermined endogenous

! ! !
variables expressed in levels (i.e., 𝑝!!! , 𝑝!!! , 𝑝!!! ) and the exogenous variables expressed in first
!
differences (i.e., Δ𝑟𝑒𝑔!!! ).19 For the Arellano and Bond estimator to yield consistent and efficient

estimates, the model’s errors cannot be autocorrelated; that is, 𝐸 𝑢!! 𝑢!! = 0 for 𝑠 ≠ 𝑡. Following

Arellano and Bond, we use the Sargan test for overidentifying restrictions, which tests the

validity of moment restrictions implied by the instruments. Under the null hypothesis that the

moment restrictions are valid (which implies the absence of second- or higher-order

autocorrelation), the test statistic is asymptotically chi-square distributed.

19
Arellano and Bond (1991) specify the use of all predetermined lagged endogenous variables, whereas we follow
the common practice of using less than the full set of lagged variables (i.e., we use periods t − 2, t − 3, and t − 4
inflation rates but not period t − 5 and prior). We did use larger instrument sets that included more lags, but the
results (not reported in this paper but available on request) were nearly identical.

17
 The Sargan test statistic for equation (5) is equal to 4.95 with an associated p value of

.176. Therefore, we cannot reject the null hypothesis that the overidentifying restrictions are

valid. Therefore, the n-step GMM estimation results reported below are both consistent and

efficient.20

The estimation results for equation (5) are given in column 2 of table 6B. Despite the

major differences in model specification and estimation of equations (4) and (5), the estimated

coefficient values are remarkably similar. Specifically, the coefficient on lagged regulatory

growth is statistically significant, equaling 0.0687, implying that a 10 percent increase in total

regulations increases consumer prices by an additional 0.687 percent.

Our results strongly support the assertion that regulatory restrictions promote inflation

across the socioeconomic spectrum, as measured by changes in the cost of baskets of goods and

services purchased by various income groups. To ensure that this result is not driven by the

basket weights themselves, we eliminate them completely and investigate the relationship

between regulatory growth and price changes for each expenditure category (e.g., bakery

products, major appliances, men’s apparel). Specifically, we estimate the following dynamic

panel model, which does not employ any household expenditure weights:

𝑝!" = 𝛼! + 𝛽𝑟𝑒𝑔!"!! + 𝜌𝑝!"!! + 𝑢!" , (6)

where 𝑝!" is the log first difference of the original price series for expenditure category 𝑖

(𝑖 = 1, … , 61), 𝛼! is the unique intercept for each expenditure category, 𝑟𝑒𝑔!"!! is the log first

difference of the regulations that apply to expenditure category 𝑖 in the prior year, and 𝑢!" is a

20
The original Arellano and Bond (1991) estimator involves two steps, whereby an initial consistent estimate of the
dynamic panel yields residuals that are used to construct a GMM weighting matrix, that is, used to more efficiently
reestimate the dynamic panel. Our software package, Eviews, iteratively repeats this process, each time updating the
GMM weighting matrix until convergence is achieved. The result is a more efficient estimator than that proposed by
Arellano and Bond.

18
mean zero error term.21 Essentially, equations (5) and (6) are very similar except that we are

modeling the price increases for individual expenditure categories rather than the broader rate of

inflation over a basket of goods. The unique intercepts accommodate different long-run rates of

inflation by category type. The results are reported in column 3 of table 6B. While smaller in

magnitude, the coefficient on lagged regulatory growth is statistically significant, equaling

0.0360, implying that a 10 percent increase in total category-specific regulatory restrictions

increases the price of goods and services in that category by an additional 0.36 percent.

Price Volatility

Clearly, increased regulations promote inflation, which is bad for all households but especially

so for poor households as they already experience the highest rate of inflation of any income

group (see table 5). Alarmingly, it is also the case that regulations are positively correlated with

price volatility. This result is especially important given the potential claim that regulations are a

form of social insurance and drive up prices but reduce price volatility. Examining the data, the

opposite is true.

For each expenditure category, we calculate its price variance and rank categories from

least to most volatile. Next, we divide the 61 categories into quartiles by volatility, with the 15

least volatile categories in quartile 1 and the 16 most volatile categories in quartile 4 (see table 7,

page 34). For each quartile, we calculate the average price variance, average price levels, average

regulations (direct, input, and total), and average budget shares for each income group over the

sample period (2000–2012). The results, provided in table 8 (page 36), are striking. Each

successive price variance quartile is much more volatile, and the average price level and average

total regulations are also sharply higher. In comparison to wealthier households, poorer

21
See table 3 for a list of the detailed expenditure categories.

19
households allocate a much larger share of total expenditures in the most volatile price

categories. In the two most stable price quartiles, wealthier households allocated 15.3 percentage

points more spending than the poorest households. By contrast, the poorest households allocated

15.3 percentage points more spending than the wealthiest households in the two most volatile

quartiles. In summary, poor households spend a substantially larger proportion of their income

on more expensive, volatile, and heavily regulated goods and services.

6. Conclusion

A significant and often hidden cost of regulation is its effect on consumer prices. As with taxes,

the burden of regulatory costs is likely to be passed along, at least in part, to consumers in the

form of higher prices. While the literature explores other specific costs of regulation, noting that

increased consumer prices are a probable consequence of heavy regulation, this study is the first

to provide a thorough empirical analysis of that relationship across industries. Our dataset, which

combines information from the CE, RegData, and price changes from the CPI, allows us to

determine the effects of regulations on prices and to ask whether those effects are regressive.

We document consumer spending patterns by income group and find that the lowest-

income households spend a larger fraction of their income in areas that are more heavily

regulated. The opposite is true of the wealthiest households; they allocate more of their spending

to goods and services that are subject to fewer regulations. Our estimates of the effect of

increased regulations on price levels suggest a positive and statistically significant relationship.

A 10 percent increase in regulations is associated with a 0.687 percent increase in prices. This

increase is particularly concerning for low-income households, which face higher levels of

overall inflation than high-income households. Finally, our analysis of price volatility suggests

that low-income households also face higher price volatility.

20
 It is important to emphasize that these results do not include state regulations. If state

regulations have a qualitatively similar impact on consumer prices, the regressive regulatory

impact of all regulations on poor households is even greater than what our results suggest. If

policymakers want to improve the welfare of the most vulnerable members of society, they

should earnestly seek ways to cut the regulatory burden faced by US firms.

21
References

Al-Ubaydli, Omar, and Patrick A. McLaughlin. 2015. “RegData: A Numerical Database on

Industry-Specific Regulations for All United States Industries and Federal Regulations,
1997–2012.” Regulation & Governance, doi: 10.1111/rego.12107.

Ardagna, Silvia, and Annamaria Lusardi. 2010. “Heterogeneity in the Effect of Regulation on
Entrepreneurship and Entry Size.” Journal of the European Economic Association 8 (2–3):
594–605.

Becker, Randy A. 2003. “Pollution Abatement Expenditure by U.S. Manufacturing Plants: Do

Community Characteristics Matter?” B.E. Journal of Economic Analysis & Policy 3 (2):
1–23.

Becker, Randy A., and J. Vernon Henderson. 2000. “Effects of Air Quality Regulations on
Polluting Industries.” Journal of Political Economy 108 (2): 379–421.

———. 2001. “Costs of Air Quality Regulation.” In Behavioral and Distributional Effects of
Environmental Policy, edited by Carlo Carraro and Gilbert E. Metcalf, 159–68. Chicago:
University of Chicago Press.

Benson, Bruce L. 2004. “Opportunities Forgone: The Unmeasurable Costs of Regulation.”

Journal of Private Enterprise 19 (Spring): 1–25.

———. 2015. “Regulation as a Barrier to Market Provision and to Innovation: The Case of Toll
Roads and Steam Carriages in England.” Journal of Private Enterprise 30 (Spring): 61–87.

Borren, Pim, and Matthew Sutton. 1992. “Are Increases in Cigarette Taxation Regressive?”
Health Economics 1 (4): 245–53.

Ciccone, Antonio, and Elias Papaioannou. 2007. “Red Tape and Delayed Entry.” Journal of the
European Economic Association 5 (2–3): 444–58.

Crafts, Nicholas. 2006. “Regulation and Productivity Performance.” Oxford Review of Economic
Policy 22 (2): 186–202.

Crain, Nicole V., and W. Mark Crain. 2010. “The Impact of Regulatory Costs on Small Firms.”
Small Business Administration, Office of Advocacy, Washington, DC, September.

Crain, W. Mark, and Nicole V. Crain. 2014. “The Cost of Federal Regulation to the U.S.
Economy, Manufacturing and Small Business.” National Association of Manufacturers,
Washington, DC, September.

Dawson, John W., and John J. Seater. 2013. “Federal Regulation and Aggregate Economic
Growth.” Journal of Economic Growth 18 (2): 137–77.

Djankov, Simeon, Caralee McLiesh, and Rita Maria Ramalho. 2006. “Regulation and Growth.”
Economics Letters 92 (3): 395–401.

22
Fisman, Raymond, and Virginia Sarria-Allende. 2004. “Regulation of Entry and the Distortion of
Industrial Organization.” Working Paper 10929, National Bureau of Economic Research,
Cambridge, MA, November.

Goldstein, Jacob, and Lam Vo. 2012. “How the Poor, the Middle Class, and the Rich Spend
Their Money.” National Public Radio, Planet Money. http://www.npr.org/sections/money
/2012/08/01/157664524/how-the-poor-the-middle-class-and-the-rich-spend-their-money.

Gørgens, Tue, Martin Paldam, and Allan Würtz. 2003. “How Does Public Regulation Affect
Growth?” Economics Working Paper 2003-14, University of Aarhus, Aarhus, Denmark.

Greenstone, Michael. 2002. “The Impacts of Environmental Regulations on Industrial Activity:

Evidence from the 1970 and 1977 Clean Air Act Amendments and the Census of
Manufactures.” Journal of Political Economy 110 (6): 1175–219.

Hazilla, Michael, and Raymond Kopp. 1990. “Social Cost of Environmental Quality
Regulations: A General Equilibrium Analysis.” Journal of Political Economy 98 (4):
853–73.

Klapper, Leora, Luc Laeven, and Raghuram Rajan. 2006. “Entry Regulation as a Barrier to
Entrepreneurship.” Journal of Financial Economics 82 (3): 591–629.

McLaughlin, Patrick A., and Richard Williams. 2014. “The Consequences of Regulatory
Accumulation and a Proposed Solution.” Working Paper 14-03, Mercatus Center, George
Mason University, Arlington, VA, February.

Miller, Sophie. 2012. “Public Interest Comment on the Department of Energy’s Direct Final
Rule: Energy Conservation Standards for Residential Dishwashers.” George Washington
University Regulatory Studies Center, Washington, DC, September.

Nicoletti, Giuseppe, and Stefano Scarpetta. 2003. “Regulation, Productivity and Growth: OECD
Evidence.” Economic Policy 18 (36): 9–72.

OMB (Office of Management and Budget). 2014. “Draft 2014 Report to Congress on the
Benefits and Costs of Federal Regulations and Unfunded Mandates on State, Local, and
Tribal Entities.” OMB, Washington, DC, May.

Poterba, James M. 1991. “Is the Gasoline Tax Regressive?” In Tax Policy and the Economy,
Volume 5, edited by David F. Bradford, 145–64. Cambridge, MA: MIT Press.

Thomas, Diana. 2012. “Regressive Effects of Regulation.” Working Paper No. 12-35, Mercatus
Center, George Mason University, Arlington, VA, November.

Wier, Mette, Katja Birr-Pedersen, Henrik Klinge Jacobsen, and Jacob Klok. 2005. “Are CO2
Taxes Regressive? Evidence from the Danish Experience.” Ecological Economics 52 (2);
239–51.

23
Table 1. Average Percentage of Total Expenditure by Income Quintiles from Consumer Expenditure Survey, 2000–2012

% Difference (Bottom Quintile

Bottom 20% Quintile 2 Quintile 3 Quintile 4 Top 20% vs. Top Quintile)
Top 25 Most-Regulated Expenditure
Categories
Direct Regulations 60.15% 58.42% 57.22% 57.22% 54.51% 10.35%
Total Regulations 58.73% 58.13% 56.64% 54.31% 52.23% 12.44%

Top 25 Least-Regulated Expenditure

Categories
Direct Regulations 25.64% 27.28% 28.99% 30.97% 32.19% -20.35%
Total Regulations 31.88% 32.22% 33.92% 36.07% 38.63% -17.47%
Source: Authors’ calculations using the Consumer Expenditure Survey and the RegData database of the Mercatus Center at George Mason University.

24
Table 2. Universal Classification Codes Categories with Largest Differences in Spending between Bottom and
Top Income Quintiles, 2000–2012

Expenditure Category Bottom Quintile Top Quintile Difference % Difference

Rented dwellings 14.67% 2.17% 12.50% 576.04%
Electricity 4.19% 2.27% 1.92% 84.58%
Medical services and insurance 5.80% 4.60% 1.20% 26.09%
Telephone services 3.25% 2.14% 1.11% 51.87%
Tobacco products and smoking supplies 1.41% 0.41% 1.00% 243.90%
Drugs and medical supplies 2.07% 1.13% 0.94% 83.19%
Miscellaneous foods 1.86% 1.26% 0.60% 47.62%
Gasoline and motor oil 4.66% 4.21% 0.45% 10.69%
Nonalcoholic beverages 1.04% 0.62% 0.42% 67.74%
Audio and visual equipment and services 2.37% 1.95% 0.42% 21.54%
Note: Numbers represent the percentage of expenditures each income quintile allocates to specific expenditure categories.
Source: Authors’ calculations using the Consumer Expenditure Survey.

25
Figure 1. Changes in Direct Regulations across Time (2000–2012) for Selected Universal Classification Codes

Note: Direct regulations, measured on the y-axes, are measured by way of industry regulation index value; see appendix A for more details.
Numbers after chart titles represent the overall percentage growth from 2000 to 2012.
Source: Authors’ calculations using the Consumer Expenditure Survey and the RegData database of the Mercatus Center at George Mason
University.

26
Table 3. Average Basket Shares and Regulations by Income Quintile

Average Basket Weights (2000 to 2012) Average Regulations

Expenditure Category Bottom 20% 2nd Quintile 3rd Quintile 4th Quintile Top 20% Direct Input Total
Alcoholic beverages 0.98% 0.96% 1.06% 1.13% 1.20% 1,000 18,139 19,140
Audio and visual equipment and services 2.37% 2.34% 2.31% 2.22% 1.95% 3,877 9,396 13,272
Bakery products 1.05% 0.96% 0.87% 0.81% 0.67% 411 15,695 16,106
Beef 0.76% 0.74% 0.65% 0.59% 0.48% 11,219 30,471 41,691
Boys' apparel, ages 2 to 15 0.23% 0.23% 0.23% 0.23% 0.24% 1,255 16,200 17,456
Cars and trucks, new 2.05% 2.72% 3.64% 4.41% 5.29% 101 6,311 6,412
Cars and trucks, used 3.55% 4.47% 4.59% 4.59% 3.33% 0 0 0
Cereals and cereal products 0.56% 0.49% 0.44% 0.39% 0.33% 180 22,418 22,597
Children's apparel, under age 2 0.26% 0.24% 0.23% 0.23% 0.18% 620 14,451 15,071
Drugs and medical supplies 2.07% 2.18% 1.71% 1.40% 1.13% 826 15,754 16,580
Education 3.12% 1.29% 1.31% 1.71% 3.30% 1,917 12,682 14,599
Eggs 0.18% 0.11% 0.11% 0.12% 0.10% 21,764 25,261 47,025
Electricity 4.19% 3.86% 3.39% 2.88% 2.27% 1,725 90,877 92,603
Fats and oils 0.34% 0.31% 0.24% 0.23% 0.16% 8 21,970 21,978
Fees and admissions 0.79% 0.81% 1.04% 1.35% 2.15% 0 29,019 29,019
Fish and seafood 0.38% 0.35% 0.31% 0.29% 0.25% 235,349 147,703 383,052
Floor coverings 0.06% 0.08% 0.09% 0.10% 0.14% 0 14,270 14,270
Food away from home 5.47% 5.61% 6.20% 6.70% 6.90% 473 15,957 16,430
Food prepared by consumer unit on out-of-town trips 0.10% 0.11% 0.11% 0.12% 0.12% 473 15,957 16,430
Footwear 0.99% 0.88% 0.84% 0.79% 0.76% 1,790 25,395 27,184
Fresh fruits 0.64% 0.60% 0.53% 0.49% 0.43% 2 17,568 17,569
Fresh milk and cream 0.53% 0.48% 0.40% 0.35% 0.25% 24 27,026 27,050
Fresh vegetables 0.62% 0.59% 0.51% 0.47% 0.41% 1 14,492 14,493
Fuel oil and other fuels 0.41% 0.39% 0.33% 0.32% 0.28% 116,284 251,824 368,108
Furniture 0.72% 0.75% 0.85% 0.95% 1.35% 17 17,310 17,327
Gasoline and motor oil 4.66% 5.33% 5.58% 5.29% 4.21% 161,726 266,598 428,323
Girls' apparel, ages 2 to 15 0.25% 0.29% 0.30% 0.31% 0.33% 1,236 16,040 17,276
Household operations 1.67% 1.81% 1.81% 2.06% 2.91% 70 6,543 6,613
Household textiles 0.25% 0.30% 0.32% 0.33% 0.37% 71 15,403 15,475
Housekeeping supplies 1.64% 1.63% 1.51% 1.61% 1.42% 9,331 22,819 32,149
Maintenance and repairs 1.62% 1.84% 1.90% 1.90% 1.79% 13,006 11,935 24,941
Major appliances 0.42% 0.47% 0.51% 0.55% 0.59% 217 13,578 13,796

27
 Average Basket Weights (2000 to 2012) Average Regulations
Expenditure Category Bottom 20% 2nd Quintile 3rd Quintile 4th Quintile Top 20% Direct Input Total
Medical services and insurance 5.80% 6.60% 6.09% 5.52% 4.60% 166,222 96,644 262,865
Men’s apparel, age 16 and over 0.69% 0.71% 0.78% 0.86% 1.00% 1,255 16,200 17,456
Miscellaneous 1.80% 1.99% 2.04% 2.04% 2.05% 34,464 19,803 54,266
Miscellaneous foods 1.86% 1.68% 1.60% 1.54% 1.26% 2 20,638 20,640
Natural gas 1.30% 1.29% 1.14% 1.03% 0.91% 18,733 259,424 278,157
Nonalcoholic beverages 1.04% 0.95% 0.86% 0.80% 0.62% 0 17,400 17,400
Other apparel products and services 0.59% 0.55% 0.54% 0.54% 0.86% 110 16,175 16,286
Other dairy products 0.69% 0.65% 0.62% 0.59% 0.51% 19 24,429 24,448
Other entertainment supplies, equipment, and services 0.57% 0.81% 0.90% 1.23% 1.44% 3,533 17,021 20,554
Other lodging 0.89% 0.76% 0.89% 1.17% 2.27% 5,352 21,054 26,406
Other meats 0.36% 0.34% 0.31% 0.25% 0.25% 20,967 30,794 51,761
Other vehicles and vehicle finance charges 0.45% 0.71% 1.03% 1.24% 0.99% 0 14,706 14,706
Owned dwellings 8.55% 10.24% 12.75% 15.50% 18.55% 84,121 51,666 135,787
Personal care products and services 1.46% 1.48% 1.45% 1.47% 1.48% 613 12,729 13,342
Pets, toys, hobbies, and playground equipment 0.96% 1.27% 1.25% 1.29% 1.31% 868 19,231 20,099
Pork 0.62% 0.55% 0.48% 0.40% 0.30% 12,844 30,525 43,369
Poultry 0.53% 0.47% 0.40% 0.38% 0.29% 11,219 39,140 50,359
Processed fruits and vegetables 0.73% 0.67% 0.55% 0.51% 0.42% 0 22,501 22,501
Public transportation 0.88% 0.85% 0.92% 1.03% 1.64% 382,599 53,333 435,932
Reading 0.30% 0.29% 0.29% 0.31% 0.33% 432 14,104 14,536
Rented dwellings 14.67% 11.38% 8.47% 4.99% 2.17% 14,741 11,343 26,084
Small appliances, misc. housewares, and household equip. 1.58% 1.72% 1.95% 2.15% 2.48% 593 16,964 17,557
Sugar and other sweets 0.41% 0.38% 0.34% 0.34% 0.27% 21 19,472 19,493
Telephone services 3.25% 3.18% 3.04% 2.74% 2.14% 33,094 13,961 47,054
Tobacco products and smoking supplies 1.41% 1.24% 1.10% 0.83% 0.41% 29,159 6,696 35,854
Vehicle insurance 2.23% 2.65% 2.75% 2.61% 2.11% 306,785 170,400 477,185
Vehicle rentals, leases, licenses, and other charges 0.75% 0.84% 1.01% 1.20% 1.53% 0 34,902 34,902
Water and other public services 1.16% 1.17% 1.09% 1.03% 0.84% 27,845 62,090 89,935
Women’s apparel, age 16 and over 1.51% 1.37% 1.43% 1.51% 1.66% 1,236 16,040 17,276

Source: Authors’ calculations using the Consumer Expenditure Survey and the RegData database of the Mercatus Center at George Mason University.

28
Table 4. Combined Household Weighted Regulations, All Households

Direct Input Total

Year Regulations Regulations Regulations
2000 42,283 41,608 83,890
2001 43,454 42,697 86,151
2002 42,998 42,661 85,659
2003 43,578 43,651 87,228
2004 45,786 46,266 92,051
2005 44,926 46,868 91,793
2006 46,056 47,990 94,046
2007 47,627 49,188 96,815
2008 50,214 53,343 103,556
2009 47,575 48,833 96,409
2010 50,569 51,759 102,328
2011 52,399 55,618 108,017
2012 54,523 57,570 112,092
Note: Regulations are measured by way of industry regulation index value; see appendix
A for details.
Source: Authors’ calculations using the Consumer Expenditure Survey and the RegData
database of the Mercatus Center at George Mason University.

29
Figure 2. Total Regulations vs. Chained Prices

120,000
Total Regul ati ons (We ighted Restrictio n Count)

110,000

100,000

90,000

80,000

70,000
90 100 110 120 130 140

Cha ined Pr ices

Note: Authors’ calculations using the RegData database of the Mercatus Center at George Mason
University and the Consumer Price Index.

30
Table 5. Laspeyres and Chained Price Indexes

Laspeyres
Year All Households Bottom 20% 2nd Quintile 3rd Quintile 4th Quintile Top 20%
2000 100 100 100 100 100 100
2001 101.114 101.388 101.216 101.149 100.999 101.117
2002 103.395 103.832 103.568 103.449 103.18 103.234
2003 104.8 105.473 105.125 104.847 104.42 104.523
2004 108.431 109.297 108.923 108.517 108.019 107.828
2005 112.241 113.488 112.967 112.342 111.663 111.272
2006 115.064 116.487 115.776 115.141 114.357 114.032
2007 120.292 122.091 121.307 120.522 119.504 118.74
2008 119.927 122.36 121.272 120.115 118.848 118.631
2009 124.303 126.703 125.765 124.819 123.432 122.479
2010 126.459 129.117 128.177 127.099 125.57 124.288
2011 130.628 133.545 132.644 131.392 129.711 128.136
2012 132.976 135.989 135.048 133.772 132.003 130.391
Inflation Rate 2.40% 2.59% 2.54% 2.45% 2.34% 2.24%

Chained
Year All Households Bottom 20% 2nd Quintile 3rd Quintile 4th Quintile Top 20%
2000 100 100 100 100 100 100
2001 100.937 101.201 101.008 100.88 100.821 100.955
2002 103.225 103.595 103.403 103.186 103.028 103.077
2003 104.479 105.248 104.827 104.35 104.001 104.257
2004 108.019 108.995 108.553 108.063 107.524 107.474
2005 111.638 113.126 112.483 111.865 111.083 110.777
2006 114.326 116.119 115.251 114.56 113.731 113.377
2007 119.122 121.388 120.529 119.649 118.631 117.646
2008 118.218 121.336 119.509 118.279 117.283 117.105
2009 122.411 125.388 124.097 122.958 121.777 120.834
2010 124.121 127.319 126.048 124.829 123.548 122.313
2011 127.872 131.422 130.034 128.741 127.312 125.842
2012 130.085 133.85 132.318 130.983 129.46 127.977
Inflation Rate 2.22% 2.46% 2.36% 2.27% 2.17% 2.08%
Source: Authors’ calculations using the Consumer Price Index and the Consumer Expenditure Survey.

31
Table 6A. Regulation Lag Length Specification and Robustness

Coefficient Equation (4): Time Series

Regulation growth (t) −0.0159 --- −0.0482 0.0087 −0.0149 --- ---
(0.0434) (0.0355) (0.0233) (0.0263)

Regulation growth (t − 1) 0.046* 0.0539** --- 0.0671*** --- 0.0648*** ---

(0.0251) (0.0226) (0.0197) (0.0213)

Regulation growth (t − 2) −0.0247 −0.0155 −0.0535 --- --- --- −0.0339

(0.0394) (0.0193) (0.0323) (0.0221)

Lagged inflation −0.4855*** −0.5106*** −0.4515*** −0.4761*** −0.4954*** −0.4651*** −0.5408***

(0.0977) (0.0405) (0.0903) (0.0599) (0.0509) (0.0366) (0.0274)

Observations 50 50 50 55 55 55 50
Goodness of Fit 0.356 0.356 0.347 0.323 0.289 0.322 0.336
Akaike Information Criterion −5.92 −5.959 −5.946 −6.018 −6.006 −6.053(a) −5.969
Schwarz Information Criterion −5.729 −5.806 −5.793 −5.872 −5.897 −5.944(a) −5.854
* = statistically significant at the 10% level; ** = statistically significant at the 5% level; *** = statistically significant at the 1% level.
(a) The baseline specification of equation (4), which includes a single, one-period lag of regulation growth, minimizes both the Akaike and Schwarz
information criteria.
Note: White robust (period) standard errors are in parentheses.

32
Table 6B. Inflation and Regulation Growth Regression Results

Equation (4) Equation (5) Equation (6)

Coefficient Time Series Dynamic Panel Expenditure Panel
Lagged regulation growth 0.0648*** 0.0687*** 0.0360***
(0.0213) (0.0148) (0.0089)

Lagged inflation −0.4651*** −0.4857*** −0.1998***

(0.0366) (0.0397) (0.0031)

Observations 55 50 610
Sargan test --- 4.95 47.36
Sargan p-value --- 0.176 0.1684
*** = statistically significant at the 1% level.
Note: White robust (period) standard errors in parentheses. Intercept for equation (4) not reported; Sargan test not
applicable to equation (4). Sargan test fails to reject null hypothesis that overidentifying restrictions are valid at any
standard level of significance in equations (5) and (6).

33
Table 7. Price Volatility of Expenditure Categories

Volatility Price Average Price Average Total

Rank Quartile Expenditure Category Price, $ Variance, $ Regulations
61 1 Audio and visual equipment and services 100.7 4.87 13,272
60 1 Cars and trucks, new 97.1 6.3 6,412
59 1 Telephone services 100.8 7.1 47,054
58 1 Women’s apparel, age 16 and over 92.6 8.43 17,276
57 1 Furniture 93.2 11.29 17,327
56 1 Footwear 100.2 12.08 27,184
55 1 Vehicle rentals, leases, licenses, and other charges 95.9 12.86 34,902
54 1 Men’s apparel, age 16 and over 91.7 14.26 17,456
53 1 Major appliances 100.5 14.68 13,796
52 1 Children's apparel, under age 2 92.5 18.73 15,071
51 1 Floor coverings 105.3 21.72 14,270
50 1 Boys' apparel, ages 2 to 15 87.6 33.52 17,456
49 1 Cars and trucks, used 88.7 37.05 0
48 1 Reading 109.8 40.02 14,536
47 1 Girls' apparel, ages 2 to 15 88.7 40.6 17,276
46 2 Housekeeping supplies 109.3 72.21 32,149
45 2 Nonalcoholic beverages 110.2 73.87 17,400
44 2 Miscellaneous foods 110.6 79.61 20,640
43 2 Personal care products and services 114.5 84.48 13,342
42 2 Fresh fruits 115.1 105.47 17,569
41 2 Pork 113.9 108.06 43,369
40 2 Owned dwellings 118.4 112.38 135,787
39 2 Fees and admissions 118.7 113.53 29,019
38 2 Other lodging 119.1 113.7 26,406
37 2 Alcoholic beverages 115.8 118.07 19,140
36 2 Other apparel products and services 103.9 120.7 16,286
35 2 Small appliances, misc. housewares, and household equip. 83.4 129.19 17,557
34 2 Other meats 118.4 135.48 51,761
33 2 Fresh vegetables 116.3 137.14 14,493
32 2 Other dairy products 116.5 138.32 24,448
31 3 Drugs and medical supplies 119.1 144.12 16,580
30 3 Cereals and cereal products 113.9 148.29 22,597
29 3 Public transportation 111.3 150.23 435,932
28 3 Food prepared by consumer unit on out-of-town trips 117.3 157.54 16,430
27 3 Poultry 118.2 160.35 50,359
26 3 Other entertainment supplies, equipment, and services 77.7 160.89 20,554
25 3 Rented dwellings 121.7 179.32 26,084
24 3 Household operations 121.3 183.93 6,613
23 3 Food away from home 119.7 190.52 16,430
22 3 Fresh milk and cream 118.8 192.91 27,050
21 3 Sugar and other sweets 116.6 193.77 19,493
20 3 Other vehicles and vehicle finance charges 115.2 226.32 14,706
19 3 Maintenance and repairs 122.5 232.8 24,941
18 3 Miscellaneous 124.3 234.73 54,266
17 3 Fish and seafood 115.6 235.39 383,052

34
 Volatility Price Average Price Average Total
Rank Quartile Expenditure Category Price, $ Variance, $ Regulations
16 4 Household textiles 75.4 239.28 15,475
15 4 Bakery products 120.8 278.52 16,106
14 4 Vehicle insurance 130.2 284.93 477,185
13 4 Processed fruits and vegetables 123.4 303.96 22,501
12 4 Pets, toys, hobbies, and playground equipment 125.8 354.58 20,099
11 4 Fats and oils 121.1 362.84 21,978
10 4 Electricity 128.4 440.12 92,603
9 4 Medical services and insurance 132.3 451.88 262,865
8 4 Natural gas 120.6 515.35 278,157
7 4 Beef 137 535.25 41,691
6 4 Eggs 129.8 613.05 47,025
5 4 Water and other public services 133.5 704.02 89,935
4 4 Education 144.3 911.07 14,599
3 4 Tobacco products and smoking supplies 151 1831.84 35,854
2 4 Gasoline and motor oil 154 3006.99 428,323
1 4 Fuel oil and other fuels 159.4 3250.45 368,108
Source: Authors’ calculations using the Consumer Expenditure Survey, the RegData database of the Mercatus
Center at George Mason University, and the Consumer Price Index.

35
Table 8. Price Volatility vs. Average Regulations and Budget Shares

Price Variance Average Price Average Price, $ Average Regulations Household Quintile Budget Shares (%) Share Differential
Quartile Variance, $ (2000 = 100) Direct Input Total Bottom Q2 Q3 Q4 Top (Top − Bottom)
1 19 96 3,009 15,211 18,219 17.4 18.8 20.1 21 20.8 3.4
2 109 112 8,997 22,961 31,958 22.3 23.5 26.1 29.5 34.2 11.9
3 186 116 46,465 29,207 75,672 31.7 29.5 27 24.5 22.5 −9.2
4 880 130 54,046 85,485 139,531 28.6 28.2 26.8 25.1 22.4 −6.1
Note: Regulations are measured by way of industry regulation index value; see appendix A for details.
Source: Authors’ calculations using the Consumer Expenditure Survey, the RegData database of the Mercatus Center at George Mason University, and the
Consumer Price Index.

36
Appendix A. Methodological Description of the Construction of the Consumer Expenditure

Survey/Regulation Dataset

To determine the disparate effects of government regulations on households in different

socioeconomic strata, we construct a dataset that maps goods and services from the Consumer

Expenditure Survey (CE) onto industry regulations from the Mercatus Center at George Mason

University’s industry regulation database (RegData).

The CE provides detailed household spending and price data for a wide array of goods

and services by income group. These goods and services are organized using the Universal

Classification Codes (UCC) system. RegData 2.0, however, reports the level of industry

regulation by the two-, three-, and four-digit North American Industry Classification System

(NAICS) code for each year between 1997 and 2012. Therefore, to construct a usable database,

we map regulations from the NAICS space onto goods and services in the UCC space. The

resulting balanced panel dataset contains 9,872 observations, covering 617 UCC-based goods

and services over a 16-year period.

To construct the final dataset, the following steps are employed:

1. The RegData 2.0 dataset consists of two-digit, three-digit, and four-digit NAICS-based

tables. Each regulation record in the tables contains the name of the government agency

imposing the regulation, the year of the regulation, the industry affected by the

regulation, the regulatory word count, the restriction count, and the industry regulation

index value. For our purposes, we use the industry regulation index value, which equals

the regulatory restriction count weighted by industry relevance.22

22
For a description of the methodology used to construct RegData, see http://regdata.org/methodology.

37
 For each industry-and-year pair, the industry regulation index values are summed

across federal regulators. Therefore, for each industry-and-year combination, a single-

industry regulation index value is derived, equaling the sum of all regulatory restrictions

(weighted by industry relevance) imposed on that industry by all federal regulators for

that year. The result is three aggregated datasets, one for each two-digit, three-digit, and

four-digit NAICS-based table. Last, the three aggregated datasets are combined (stacked)

to form a single dataset.

2. The spreadsheet containing the 2007 commodity-by-industry direct requirements (after

redefinitions) table was downloaded from the Bureau of Economic Analysis (BEA)

website (http://www.bea.gov/industry/xls/io-annual/CxI_DR_2007_detail.xlsx). This

spreadsheet contains two work sheets, both of which are used below:

a. The first work sheet is a concordance that converts the BEA’s input-output (I-O)

commodity/industry codes into 2007 NAICS codes.

b. The second work sheet is the I-O direct requirements table, which contains I-O

weights (𝛼!" ) equal to the amount of input (measured in dollars) from industry

(𝑖 ) required to produce a dollar’s worth of output by industry (𝑗). By

construction, these weights sum to 1 because, in addition to actual inputs, the

BEA includes employee compensation, taxes, and gross operating surplus in the

weighting schema.

3. The I-O commodity/industry code to NAICS concordance described in step (2a) above

is matched with the aggregate industry regulations from step (1), to create a new table

that lists the aggregate industry regulations by I-O commodity/industry code; the

resulting table is further summed over commodity code by year to derive a table with a

38
 single total regulation value for each commodity code–year pair. This second round of

aggregation after the initial match is necessary because some commodity codes map

onto multiple NAICS industries. I-O commodity/industry codes with no associated

regulations are assigned an industry regulation index value of 0. The resulting table is a

measure of the direct regulations (denoted 𝐷𝑖𝑟𝑒𝑐𝑡𝑅𝑒𝑔!" ) applicable to a given I-O

commodity/industry code.

4. To determine the level of regulation that applies to the inputs/supply chain of a given

industry, the I-O direct requirements (𝛼!" ) from step (2b) are matched with the direct

regulations for each I-O commodity (𝐷𝑖𝑟𝑒𝑐𝑡𝑅𝑒𝑔!" ) from step (3) by way of their I-O

commodity/industry codes. Note that if a commodity/industry is not needed to produce a

given output, the associated input value is 0. This produces a large result set with more

than 2.4 million rows of data. This dataset is then “grouped by” output industry (𝑗) and

year (𝑡) and summed over the product of the direct input regulations (indexed by i) and I-

O weights, producing an estimate of input–supply chain regulation:

𝐼𝑛𝑝𝑢𝑡𝑅𝑒𝑔!" = ! 𝛼!" ∙ 𝐷𝑖𝑟𝑒𝑐𝑡𝑅𝑒𝑔!" .

See figure A1 (page 42) for a graphical summary of steps (1) to (4).

5. The direct regulations by industry and year are matched with the total input regulations

by industry and year. The direct and input regulations are summed to determine the total

direct and indirect regulations affecting a given industry:

𝑇𝑜𝑡𝑎𝑙𝑅𝑒𝑔!" = 𝐷𝑖𝑟𝑒𝑐𝑡𝑅𝑒𝑔!" + 𝐼𝑛𝑝𝑢𝑡𝑅𝑒𝑔!" .

6. To map regulations onto the UCC codes, a separate set of queries is executed to map the

codes onto I-O commodity/industry codes.

39
 a. As a beginning step, we import the personal consumption expenditures (PCE)

concordance from the Bureau of Labor Statistics (BLS) (http://www.bls.gov/cex

/pce_concordance_2012.xlsx). This file maps UCC codes onto PCE codes from

the BEA’s national income and product accounts (NIPAs).

b. Next, we import BEA table 2.4.5U (I-O, Personal Consumption Expenditures by

Type of Product with 2007 Input-Output Commodity Composition). This latter

bridge file (http://www.bea.gov/national/xls/2007-pcs-io-bridge.xls) maps NIPA

line numbers onto PCE codes.

c. The tables from steps (6a) and (6b) are matched by way of their common PCE

codes. The resulting table serves as a bridge file that maps UCC codes onto NIPA

line numbers.

7. Finally, we import the BEA’s PCE bridge file, which maps NIPA line numbers onto I-O

commodity/industry codes (www.bea.gov/industry/xls/io-annual/PCEBridge_2007

_Detail.xlsx), along with the total value of all purchases of the linked I-O

commodity/industry in 2007.

a. Matching the NIPA line items from the PCE bridge with the results from step (6c)

provides a clear mapping from UCC code to I-O commodity/industry codes. See

figure A2 (page 43) for a graphic summary of steps (6) and (7).

8. The resulting table from step (7a) maps a given consumer product from the CE onto all I-

O industries that produce that product. In many cases, more than one industry produces a

given UCC product. To produce a single regulation value for each consumer product, we

derive industry weights equal to a given industry’s 2007 level of output relative to the

40
 total output of all industries that supply a given UCC product.23 For example, the UCC

code for flour is 10110. This consumer product is produced by seven I-O industries.

Assigning each of these industries a weight equal to its total output relative to the total

output of all seven industries produces a set of weights that sum to 1 (see table A1, page

44). Although it would be preferable to update these weights annually, the BLS derives

these output data from the US Census Bureau’s Economic Census, which is conducted

only every five years.

9. Finally, UCC codes, I-O commodity/industry codes, and output shares from step (8) are

matched with the regulation-by-industry data from step (5). These matched data are then

“grouped by” UCC code and year and aggregated over the product of industry regulation

and output shares.

23
Consumption-based weights equal to each industry’s market share for a given commodity would be preferable to
weights based on the overall relative size of the industries that produce said commodity. Unfortunately, to our
knowledge, such data do not exist.

41
Figure A1. Mapping Regulations onto Input-Output (I-O) Codes

Aggregated RegData NAICS Codes Concordance

Source: Mercatus Center Source: Bureau of Economic Analysis
Appendix A step (1) Appendix A step (2a)
NAICS code NAICS code
Total regulation index I-O commodity/industry code

I-O Commodity-by-Industry Regulations by I-O Code

Direct Requirements Table Source: Combine steps (1) and (2a) above
Source: Bureau of Economic Analysis Appendix A step (3)
Appendix A step (2b)

I-O commodity/industry code I-O commodity/industry code

I-O direct requirement coefficient (αij) Total regulation index (DirectRegit)

Total Direct Regulations by I-O Code

Source: Combine steps (2b) and (3) above
Appendix A step (4)

I-O commodity/industry code

InputRegjt = Σ (αij × DirectRegit)

Note: NAICS = North American Industry Classification System.

42
Figure A2. Mapping Input-Output (I-O) Codes onto Consumer Expenditure Codes

PCE Concordance I-O, Personal Consumption

Source: Bureau of Labor Statistics Expenditures by Type of Product
Appendix A step (6a) Source: Bureau of Economic Analysis
Appendix A step (6b)
PCE code PCE code
UCC code NIPA line number

PCE Bridge UCC Code by NIPA Line Number

Source: Bureau of Economic Analysis Source: Combine steps (6a) and (6b) above
Appendix A step (7) Appendix A step (6c)

NIPA line number NIPA line number

I-O commodity/industry code UCC code

UCC Code by I-O Code

Source: Combine steps (6c) and (7) above
Appendix A step (7a)

I-O commodity/industry code

UCC code

Note: PCE = personal consumption expenditures; UCC = Universal Classification Codes; NIPA = national income
and products accounts.

43
Table A1. Input-Output Industries that Produce Flour (UCC: 10110)

Commodity Purchase Output

Code Commodity/Industry Description Value Share, %
311230 Breakfast cereal manufacturing 12,889 34.7
31122A Soybean and other oilseed processing 114 0.3
3118A0 Cookie, cracker, pasta, and tortilla manufacturing 16,255 43.8
311210 Flour milling and malt manufacturing 4,659 12.5
311990 All other food manufacturing 660 1.8
1111B0 Grain farming 618 1.7
311420 Fruit and vegetable canning, pickling, and drying 1,939 5.2
Source: Authors’ calculations using the Consumer Expenditure Survey, the Bureau of Economic Analysis’s PCE
bridge file, and the Bureau of Labor Statistics’s PCE concordance file.

44
Appendix B. Top 20 Expenditure Categories by Income Quintile and

Corresponding Regulations

Income Quintile 1 (Bottom 20%)

Expenditure Category % Expenditure Direct Reg Rank Direct Regs Total Reg Rank Total Regs
Rented dwellings 14.67% 15 14,741 25 26,084
Owned dwellings 8.55% 7 84,121 8 135,787
Medical services and insurance 5.80% 4 166,222 7 262,865
Food away from home 5.47% 37 473 45 16,430
Gasoline and motor oil 4.66% 5 161,726 3 428,323
Electricity 4.19% 26 1,725 9 92,603
Cars and trucks, used 3.55% 55 0 61 0
Telephone services 3.25% 9 33,094 14 47,054
Education 3.12% 24 1,917 52 14,599
Audio and visual equipment and services 2.37% 22 3,877 58 13,272
Vehicle insurance 2.23% 2 306,785 1 477,185
Drugs and medical supplies 2.07% 33 826 44 16,580
Cars and trucks, new 2.05% 44 101 60 6,412
Miscellaneous foods 1.86% 53 2 31 20,640
Miscellaneous 1.80% 8 34,464 11 54,266
Household operations 1.67% 46 70 59 6,613
Housekeeping supplies 1.64% 20 9,331 20 32,149
Maintenance and repairs 1.62% 16 13,006 26 24,941
Small appliances, misc. housewares, and household equip. 1.58% 36 593 37 17,557
Women’s apparel, age 16 and over 1.51% 30 1,236 43 17,276

Income Quintile 2
Expenditure Category % Expenditure Direct Reg Rank Direct Regs Total Reg Rank Total Regs
Rented dwellings 11.38% 15 14,741 25 26,084
Owned dwellings 10.24% 7 84,121 8 135,787
Medical services and insurance 6.60% 4 166,222 7 262,865
Food away from home 5.61% 37 473 45 16,430
Gasoline and motor oil 5.33% 5 161,726 3 428,323
Cars and trucks, used 4.47% 55 0 61 0
Electricity 3.86% 26 1,725 9 92,603
Telephone services 3.18% 9 33,094 14 47,054
Cars and trucks, new 2.72% 44 101 60 6,412
Vehicle insurance 2.65% 2 306,785 1 477,185
Audio and visual equipment and services 2.34% 22 3,877 58 13,272
Drugs and medical supplies 2.18% 33 826 44 16,580
Miscellaneous 1.99% 8 34,464 11 54,266
Maintenance and repairs 1.84% 16 13,006 26 24,941
Household operations 1.81% 46 70 59 6,613
Small appliances, misc. housewares, and household equip. 1.72% 36 593 37 17,557
Miscellaneous foods 1.68% 53 2 31 20,640
Housekeeping supplies 1.63% 20 9,331 20 32,149
Personal care products and services 1.48% 35 613 57 13,342
Women’s apparel, age 16 and over 1.37% 30 1,236 43 17,276
Income Quintile 3
Expenditure Category % Expenditure Direct Reg Rank Direct Regs Total Reg Rank Total Regs
Owned dwellings 12.75% 7 84,121 8 135,787
Rented dwellings 8.47% 15 14,741 25 26,084
Food away from home 6.20% 37 473 45 16,430
Medical services and insurance 6.09% 4 166,222 7 262,865
Gasoline and motor oil 5.58% 5 161,726 3 428,323
Cars and trucks, used 4.59% 55 0 61 0
Cars and trucks, new 3.64% 44 101 60 6,412
Electricity 3.39% 26 1,725 9 92,603
Telephone services 3.04% 9 33,094 14 47,054
Vehicle insurance 2.75% 2 306,785 1 477,185
Audio and visual equipment and services 2.31% 22 3,877 58 13,272
Miscellaneous 2.04% 8 34,464 11 54,266
Small appliances, misc. housewares, and household equip. 1.95% 36 593 37 17,557
Maintenance and repairs 1.90% 16 13,006 26 24,941
Household operations 1.81% 46 70 59 6,613
Drugs and medical supplies 1.71% 33 826 44 16,580
Miscellaneous foods 1.60% 53 2 31 20,640
Housekeeping supplies 1.51% 20 9,331 20 32,149
Personal care products and services 1.45% 35 613 57 13,342
Women’s apparel, age 16 and over 1.43% 30 1,236 43 17,276

Income Quintile 4
Expenditure Category % Expenditure Direct Reg Rank Direct Regs Total Reg Rank Total Regs
Owned dwellings 15.50% 7 84,121 8 135,787
Food away from home 6.70% 37 473 45 16,430
Medical services and insurance 5.52% 4 166,222 7 262,865
Gasoline and motor oil 5.29% 5 161,726 3 428,323
Rented dwellings 4.99% 15 14,741 25 26,084
Cars and trucks, used 4.59% 55 0 61 0
Cars and trucks, new 4.41% 44 101 60 6,412
Electricity 2.88% 26 1,725 9 92,603
Telephone services 2.74% 9 33,094 14 47,054
Vehicle insurance 2.61% 2 306,785 1 477,185
Audio and visual equipment and services 2.22% 22 3,877 58 13,272
Small appliances, misc. housewares, and household equip. 2.15% 36 593 37 17,557
Household operations 2.06% 46 70 59 6,613
Miscellaneous 2.04% 8 34,464 11 54,266
Maintenance and repairs 1.90% 16 13,006 26 24,941
Education 1.71% 24 1,917 52 14,599
Housekeeping supplies 1.61% 20 9,331 20 32,149
Miscellaneous foods 1.54% 53 2 31 20,640
Women’s apparel, age 16 and over 1.51% 30 1,236 43 17,276
Personal care products and services 1.47% 35 613 57 13,342

46
Income Quintile 5
Expenditure Category % Expenditure Direct Reg Rank Direct Regs Total Reg Rank Total Regs
Owned dwellings 18.55% 7 84,121 8 135,787
Food away from home 6.90% 37 473 45 16,430
Cars and trucks, new 5.29% 44 101 60 6,412
Medical services and insurance 4.60% 4 166,222 7 262,865
Gasoline and motor oil 4.21% 5 161,726 3 428,323
Cars and trucks, used 3.33% 55 0 61 0
Education 3.30% 24 1,917 52 14,599
Household operations 2.91% 46 70 59 6,613
Small appliances, misc. housewares, and household equip. 2.48% 36 593 37 17,557
Electricity 2.27% 26 1,725 9 92,603
Other lodging 2.27% 21 5,352 24 26,406
Rented dwellings 2.17% 15 14,741 25 26,084
Fees and admissions 2.15% 55 0 21 29,019
Telephone services 2.14% 9 33,094 14 47,054
Vehicle insurance 2.11% 2 306,785 1 477,185
Miscellaneous 2.05% 8 34,464 11 54,266
Audio and visual equipment and services 1.95% 22 3,877 58 13,272
Maintenance and repairs 1.79% 16 13,006 26 24,941
Women’s apparel, age 16 and over 1.66% 30 1,236 43 17,276
Public transportation 1.64% 1 382,599 2 435,932
Note: Regulations are measured by way of industry regulation index value; see appendix A for details.
Source: Authors’ calculations using the Consumer Expenditure Survey and the RegData database of the Mercatus
Center at George Mason University.

47