Antitrust Enforcement During the Great Depression Jenny R. Hawkins * Preliminary and Incomplete - Please Do Not Circulate May 2014 Abstract: We offer an empirical study of antitrust enforcement by the Department of Justice (DOJ) and Federal Trade Commission (FTC) during the New Deal. The goal of this research is to address arguments that antitrust policies during the Great Depression were harmfully lax or even “suspended”. These arguments stem from the National Industrial Recovery Act (NIRA), lasting from 1933 to 1935, that allowed industries to collude in specific ways under specific rules. Therefore, on the surface, the NIRA indicates antitrust policies were nonexistent during this period. Other non-legislative antitrust actions by regulators and the courts also affected antitrust policy. How- ever, changes in antitrust policy do not imply no enforcement of competition law during this period. We seek to resolve this growing assumption in the literature by creating a detailed dataset, and then carefully examining federal antitrust cases from 1925-1939. Resolving this growing assumption of nonexistent antitrust enforcement is important because some base conclusions on the extent of the Great Depression and output on an assumption of suspended antitrust policy. Such policy im- plications during this period have been paralleled to potential policy changes in today’s economic environment. Our analysis provides a detailed look at the industries and alleged violations inves- tigated and prosecuted by the DOJ and FTC, as well as other non-legislative antitrust policies in place, the size of the regulatory agencies, the resulting remedies, and length of cases. We examine the years 1925-1939 due to vacillating antitrust enforcement perspectives of the government between these years, as well as to compare to periods outside of the NIRA. Complementing our case data with specifics of the NIRA Codes of Fair Competition (rules of collusion for each industry), as well as data on other non-legislative antitrust policy, we determine factors that explain antitrust enforce- ment, thereby providing a clearer picture of exactly how antitrust enforcement responded to changes in antitrust policy during this fragile period. As of the date of this draft, data is still being entered and we have not examined any data in detail. * Case Western Reserve University, Department of Economics. I thank Price Fishback and Omar Farooque for helpful discussions and suggestions. I also thank Christopher Danis for assistance entering data.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Antitrust Enforcement During the New Deal?Jenny R. Hawkins∗
May 2014
Abstract: We offer an empirical study of antitrust enforcement by
the Department of Justice
(DOJ) and Federal Trade Commission (FTC) during the New Deal. The
goal of this research is to
address arguments that antitrust policies during the Great
Depression were harmfully lax or even
“suspended”. These arguments stem from the National Industrial
Recovery Act (NIRA), lasting
from 1933 to 1935, that allowed industries to collude in specific
ways under specific rules. Therefore,
on the surface, the NIRA indicates antitrust policies were
nonexistent during this period. Other
non-legislative antitrust actions by regulators and the courts also
affected antitrust policy. How-
ever, changes in antitrust policy do not imply no enforcement of
competition law during this period.
We seek to resolve this growing assumption in the literature by
creating a detailed dataset, and
then carefully examining federal antitrust cases from 1925-1939.
Resolving this growing assumption
of nonexistent antitrust enforcement is important because some base
conclusions on the extent of
the Great Depression and output on an assumption of suspended
antitrust policy. Such policy im-
plications during this period have been paralleled to potential
policy changes in today’s economic
environment. Our analysis provides a detailed look at the
industries and alleged violations inves-
tigated and prosecuted by the DOJ and FTC, as well as other
non-legislative antitrust policies in
place, the size of the regulatory agencies, the resulting remedies,
and length of cases. We examine
the years 1925-1939 due to vacillating antitrust enforcement
perspectives of the government between
these years, as well as to compare to periods outside of the NIRA.
Complementing our case data
with specifics of the NIRA Codes of Fair Competition (rules of
collusion for each industry), as well
as data on other non-legislative antitrust policy, we determine
factors that explain antitrust enforce-
ment, thereby providing a clearer picture of exactly how antitrust
enforcement responded to changes
in antitrust policy during this fragile period. As of the date of
this draft, data is still being entered
and we have not examined any data in detail.
∗Case Western Reserve University, Department of Economics. I thank
Price Fishback and Omar Farooque for helpful discussions and
suggestions. I also thank Christopher Danis for assistance entering
data.
1 Introduction
In June 1933, the National Industrial Recovery Act (NIRA) was
passed under emergency
legislation in an effort to fight the rapid decline in prices and
wages in the early years of
the Great Depression. While the NIRA was limited to two years (with
intent of renewal),
the goals of the legislation went beyond the Great Depression in a
permanent way. In the
words of President Franklin D. Roosevelt, “it represents a supreme
effort to stabilize for
all time the many factors which make for the prosperity of the
nation and the preservation
of American standards.” (Dearing, Homan, Lorwin, and Lyon (1934)).
Not only was the
goal grand, but too were the provisions of the NIRA. In simple
terms, Title I of the NIRA
allowed trade practices that aimed to encourage production and
employment to return to
“normal” levels (Nathan (1935)). Under specific terms, it allowed
for government-supported
collusion of firms within approved industries. Before the NIRA
could be renewed, in May
1935, the United States Supreme Court (“Court”) declared it
unconstitutional. The two
years of the NIRA, coupled with (i) government encouragement of
trade associations and
industry codes and (ii) the Court’s ruling in the price fixing case
Appalachian Coals, Inc.
v. United States (1933)1 certainly changed antitrust policy in the
United States during this
period (Bittlingmayer (1995)).
One important question, and the relevancy to today’s economic
situation, is whether,
contrary to the government’s intentions, these changes in antitrust
policy during this period
negatively affected economic recovery. Cole and Ohanian (2004) and
Eggertsson (2012)
both address this question in considering a dynamic general
equilibrium model of the NIRA.
Cole and Ohanian (2004) conclude that under the assumption of
monopoly pricing during
the Great Depression, the contraction lasted seven years longer
than it may have otherwise.2
While both of the Cole and Ohanian (2004) and Eggertsson (2012)
models hinge on an
1The Court’s decision in Appalachian Coals changed the per se rule
in price fixing to a rule of reason. The per se rule was reinstated
in 1940 in U.S. v. Socony-Vacuum Oil (310 U.S. 150 [1940].
2This conclusion of Cole and Ohanian (2004) was popularized in the
media amidst several discussions of their research, including
articles such as “How Government Prolonged the Depression” February
2, 2009, The Wall Street Journal Opinion page:
http://online.wsj.com/article/SB123353276749137485.html#.
assumption that firms collude3, Eggertsson (2012) finds that output
actually increases,
in favor of these policies. However, despite changes in antitrust
policy, it remains to be
determined the correctness of the assumption made by both Cole and
Ohanian (2004) and
Eggertsson (2012) that collusion was allowed to run rampant and
antitrust policies were
essentially non existent during the New Deal. In other words, a
change in antitrust policy
towards more lenient measures does not necessarily imply antitrust
enforcement during this
period was lax or nonexistent.
The assumptions made by Cole and Ohanian (2004) and Eggertsson
(2012) are not
uncommon in the literature; others claim that antitrust policies
were not enforced during
the New Deal. While Bittlingmayer (1995) offers a thoughtful look
at the effects of the NIRA
on recovery, he refers to the NIRA as “suspension of antitrust”.
However, his work indicates
this statement to be a misleading generalization, common in the
literature. He continues
to indicate that only several major industries participated in the
NIRA by government-
approved restriction of output or capacity, collusion on pricing or
costs information, of
prohibition of sales below costs, as is evident in studies of the
NIRA. Regardless, even
if only a small subset of industries participated in
government-sponsored collusion, it is
unclear whether these industries still behaved in
non-government-approved anticompetitive
behavior probed by regulators. Cox (1981) states, “Even though
there was little evidence of
an increase in monopoly [during the Great Depression], the belief
was widely held,” and he
provides an example from Berle and Means (1932) who argue that
industries had become
highly concentrated, yet their evidence does not support this
claim.
In an effort to clear up the enforcement of antitrust policies
during the New Deal, this
research addresses (i) the extent to which antitrust policies
during the New Deal affected
antitrust enforcement during the Great Depression and (ii) whether,
outside the limitations
of the NIRA and Appalachian Coals, antitrust regulators were
otherwise lax in enforcement.
The first goal of this research is to carefully investigate
antitrust enforcement during
the New Deal to understand exactly how antitrust policies affected
enforcement. Most of
3 Eggertsson (2012) states that “The NIRA declared a temporary
“emergency” that suspended antitrust laws...”.
3
the literature focuses on evaluation of the NIRA and its codes to
examine collusion (see,
e.g., Nathan (1935), Taylor (2007), Alexander (1997) and Alexander
and Libecab (2000)).
Posner (1970) is the only work to take a substantial look at
antitrust enforcement, though his
focus is not on the Great Depression, and his analysis does not
consider the NIRA codes nor
does he present a yearly analysis of the industries and
goods/services for which regulators
enforced antitrust policy. In the spirit of Posner (1970), we look
at DOJ and the FTC cases
between the years 1925-1939 to examine which industries,
goods/services and antitrust
violations were the focus of regulators during these years to gain
a better understanding
of antitrust enforcement under Franklin D. Roosevelt’s leadership.
We also consider the
length of cases, sanctions and remedies, types of violators (firms
versus individuals), DOJ
and FTC antitrust appropriations, and civil versus criminal pursuit
of enforcement to clear
up the picture of antitrust policy beyond the NIRA.
Then, incorporating (i) the industries and codes of collusion using
data from the NIRA’s
Codes of Fair Competition and (ii) non-legislative data regarding
influential Supreme Court
antitrust cases during this time and the FTC’s encouragement and
sponsorship of industry
codes and trade association conferences , we hope to combine our
DOJ and FTC case
information to more carefully examine enforcement of the antitrust
policies during this
period, even when government policies allowed for some violations
of antitrust laws.
2 Antitrust policy through the mid 1930s
With fairly young antitrust policy in place at this time, part of
our investigation of
antitrust enforcement during the New Deal includes a more detailed
look at the policies
and the regulators dictating the law, outlined in Table 1, below.
Antitrust policy was
led primarily by the Sherman Act of 1890 and the Clayton Act of
1914 (along with its
amendment, the Robinson-Patman Act of 1936).
After passage of the Sherman Act in 1890 and up to 1933, the
Assistant to the Attorney
General dealt with antitrust matters. In 1933 the Antitrust
Division of the Department
4
of Justice was established under Franklin D. Roosevelt, with Harold
M. Stevens appointed
as the first Assistant Attorney General of the Antitrust Division.
Along with the DOJ’s
Antitrust Division, the FTC (as part of the Federal Trade
Commission Act of 1914) also
regulates competition law in the United States. The Robinson-Patman
Act, which focuses
on price discrimination, is enforced primarily by the FTC and not
the DOJ.
A closer look at the focus of various antitrust regulators also can
reveal the relative
enforcement of antitrust policies. Following Theodore Roosevelt’s
strict enforcement of
antitrust, during the teens and through the 1920s, antitrust
policies were somewhat lax
through the allowance of trade associations and even sponsored
“Trade Practice Confer-
ences” by the FTC beginning in 1925. As a regulator, Herbert Hoover
supported trade as-
sociations; however, as President, he was a strict enforcer of
antitrust policies and rejected
any proposals to relaxed enforcement. On October 25, 1929, the
Hoover administration
took a public stance by declaring strict enforcement of antitrust
policy. This stemmed from
the argument that lax antitrust policies had created an unhealthy
boom in mergers and the
stock market during the 1920s (Bittlingmayer (1995)). Once Franklin
D. Roosevelt took
office and signed the NIRA in 1933, he had loosened the policies of
Hoover back to the
aims of the Coolidge administration for promoting exchange of
information on prices, costs
and production through trade associations (also known as the “Open
Price Movement”).
However, Roosevelt’s administration was not particular to lax
antitrust enforcement. In
1938, President Franklin D. Roosevelt appointed Thurman Arnold as
head of the Antitrust
Division of the DOJ. Arnold became known as one of the staunchest
antitrust enforcers to
date.
As Hawley (1966) points out, this alternating enforcement of
antitrust laws began with
the Sherman Act and continued through the Great Depression as
different administrations
fought what they believed was “cutthroat competition”. However
despite this seesawing
policy, it appears that regardless of the administration in office,
the FTC was constant
in its support of first, trade associations, then trade association
conferences, and finally
proposed industry codes of industry behavior. These industry codes
were supported by the
5
FTC under the belief it was their obligation to determine and then
sanction such behavior.
Yet, arguably these conferences and codes masked collusion by
firms.
The DOJ and U.S. Supreme Court also exhibited some decisions in the
1920s that
indicate some lax enforcement of antitrust policy during those
years. With the appointment
of William Donovan as Assistant to the Attorney General, and
therefore head of antitrust,
the DOJ appeared to advise trade associations, though it is unclear
to what extent and
in what capacity. Several important Supreme Court cases also
loosened the precedent on
mergers, first in U.S. vs U.S. Steel (1920), where U.S. Steel
gained 90 percent of the steel
industry’s market share, and then in U.S. vs International
Harvester (1927). However,
by the Hoover administration, the DOJ was strictly enforcing
antitrust laws against trade
associations that had formed industry codes supported by the FTC
(and possibly the DOJ
itself) in the mid-1920s. It also required a change in the FTC’s
Trade Practice Conferences
and the types of codes created therewith (Bittlingmayer
(1995)).
Besides the Open Price Movement and its continued promotion through
Trade Prac-
tice Conferences, the NIRA was the other major antitrust policy
during this period. As
mentioned above, the NIRA was passed in June 1933 for a two year
period, but before it
could be renewed, the Supreme Court declared it unconstitutional in
1935. In an attempt
to combat the collapse in output during 1929-1933, the National
Recovery Administra-
tion (NRA), the regulatory agency created to oversee the NIRA,
approved “Codes of Fair
Competition” proposed by trade associations. With approval, the
codes held for the en-
tire industry pertaining to those codes. All of the codes were
exempt from antitrust law.
The President was given the power to prescribe new codes, as well
as cancel or modify any
previously established codes. The legislation further allowed for
“agreements” between peo-
ple in the industry, trade associations, and labor associations
that also were exempt from
antitrust laws (Nathan (1935). Thus far, we have no evidence of the
extent of such agree-
ments. The NIRA also provided for collective bargaining and the
freedom to join unions, as
well as requiring employers to adhere to restrictions on minimum
pay and maximum hours
6
(Bittlingmayer (1995)).4
During the years the NIRA was in effect, almost 800 industry codes
were passed. Among
those industries not participating were agricultural, steam
railroads, nonprofit institutions,
and professional services (Alexander (1997)). According to Taylor
(2007), larger industries
were favored in faster and possibly more lenient approval of codes.
Bittlingmayer (1995)
offers evidence that while some major industry codes were detailed
and clearly anticompeti-
tive in the usual sense, other major industries did not focus on
price collusion. For example,
the automobile code focused on labor issues. In our analysis, a
closer look at the industries
for which regulators prosecuted firms and the codes existing for
those industries may reveal
more information about antitrust enforcement. Additionally,
following the NIRA, some
industries still adhered to previously signed codes. One would
expect tougher antitrust
authorities to therefore begin tighter regulation on such
industries openly continuing to
collude.
Our goal is to take the known changes in antitrust policy,
specifically FTC Trade Prac-
tice Conferences and initial creation of industry codes, the NIRA,
and changed precedence
in mergers and price fixing through Supreme Court decisions to then
examine the anticom-
petitive violations of firms prosecuted by the DOJ and FTC in order
to determine how
antitrust enforcement ensued during this turbulent period.
Antitrust policy following the NIRA is less clear. As indicated
above, at least some
industries continued to collude, and enforcement of antitrust
policies within firms in these
industries is to be determined in our examination.
3 Related Literature
As mentioned above, Cole and Ohanian (2004), Eggertsson (2012), and
Bittlingmayer
(1995) seek to answer whether output increased or decreased due to
the passage of the
NIRA. While this is not our focus, these questions point out the
important question of
4These regulations covered Title I of the NIRA, with Title II
addressing public works/construction projects and Title III
addressing provision related to amendments of the Emergency Relief
and Construction Act.
7
1890 Sherman Antitrust Act
1901-1909 Strict antitrust enforcement under Theodore
Roosevelt
1910-1920 Open Price Movement where trade associations sought to
ex- change information freely
1914 Clayton Act
1920 Leniency by SCOTUS in mergers: U.S. vs U.S. Steel (1920)
decided
1923-1929 Lenient antitrust enforcement under Calvin Coolidge
(promo- tion of trade associations)
1924 Continued leniency under William J. Donovan, appointed DOJ
Assistant to the Attorney General
1925 FTC begins sponsoring “Trade Practice Conferences”
1927 Leniency by SCOTUS in mergers: U.S. vs International Har-
vester (1927) decided
1929-1933 Strict antitrust enforcement under Herbert Hoover
1929 Numerous industries have adopted “codes” under FTC guid-
ance
1929 (Oct 25) Hoover declares strict enforcement of antitrust
1929 (Oct 29) Black Tuesday
1933 (March) Appalachian Coals, Inc. vs United States decided by
SCOTUS making price fixing easier
1933 (June) NIRA Passed
1933 (July 16) FDR signs first code (for cotton textiles)
1933 (July 19-Aug 31) President’s Reemployment Agreement (“blanket
code”) signed for labor issues
1934 Compliance Crisis
1935 (May 27) NIRA declared unconstitutional (via Schechter Poultry
Corp. vs U.S.
1935 (summer) 90 percent of cotton textile firms still adhering to
industry codes (though not legally enforceable)
1936 Robinson-Patman Act on price discrimination
1938 Thurman Arnold appointed Assistant Attorney General of the
Antitrust Division
8
how strict antitrust enforcement was during the New Deal, holding
constant the change in
antitrust policy. Others look at the NIRA and industry codes to
examine how industries
actually responded to the NIRA (see e.g., Taylor (2007), Alexander
and Libecab (2000),
and Alexander (1997)). Cox (1981) is the only known work that
attempts to investigate
monopoly behavior before, during and after the Great Depression.
However, his focus is
not on the effects of antitrust policy and antitrust enforcement,
but the ex post realization
of monopolies.
Posner (1970) not only enumerates DOJ (and certain, but not all
FTC) antitrust cases
from 1890-1969, but groups them in years of five by violation. This
enumeration, while
useful in a grand scheme, is not as useful for a detailed analysis
of antitrust policies during
the New Deal. Therefore, in the spirit of Posner, we look at DOJ
and FTC antitrust cases
during 1925-1939 on a yearly basis and break them down by violation
and industry, as well
as look at other factors neglected by Posner (1970) such as budget
appropriations and FTC
trade conferences. Additionally, Posner does not offer a detailed
statistical analysis; he
provides no statistical tests nor the relationships between
variables.
Posner (1970) provides some theoretical discussion regarding what
determines antitrust
enforcement. With respect to an economic downturn, there’s more
incentive to violate an-
titrust laws during downturns, but there may be fewer resources to
reveal and investigate
such violations. Therefore, does the observation of fewer
investigations indicate fewer viola-
tions or less enforcement? Posner (1970) suggests two hypotheses
related to economic activ-
ity: (i) Level of economic activity and antitrust enforcement are
positively correlated, as dur-
ing upturns, there are more resources for enforcement, even if
there is less incentive to violate
antitrust laws. (ii) Level of economic activity and antitrust
enforcement are negatively corre-
lated. During economic contractions, the government may be harsher
on anticompetitive be-
havior, seeking violations more vigorously. As Gallo, Dau-Schmidt,
Craycraft, and Parker
(2000) indicate, this follows the theory of regulation by Peltzman
(1976) where“regulation
will tend to be more heavily weighted toward ‘producer protection’
in depressions and
toward ‘consumer protection’ in expansion”. This theory might seem
to apply during the
9
Great Depression and with respect to the NIRA, as firms have more
incentive to seek protec-
tion from the government during downturns due to increased
competition resulting from ex-
cess capacity experienced during a contraction. Gallo, Dau-Schmidt,
Craycraft, and Parker
(2000) suggest this behavior was observed during the 1930s, but
does evidence show the
government was less stringent on firms during the Great Depression,
when controlling for
the size of the DOJ and FTC and other factors?
Posner (1970) finds a positive correlation between the number of
DOJ antitrust cases
and real GNP for the years 1890-1940, generally. From his Figure 1,
we see that pre-Great
Depression, correlation is unclear; as GNP slowly increases, cases
widely fluctuated and
decrease at a point of an increased growth rate between 1925-1030.
During the Great
Depression, the trend in cases appears to follow the trend in GNP.
Posner’s analysis for the
years 1940-1970 reveals that even though real GNP rose, the number
of antitrust cases did
not increase significantly. Gallo, Dau-Schmidt, Craycraft, and
Parker (2000) find similar
results from 1955-1972 and an even stronger case against the
positive correlation theory in
the 1980s, as they find that the number of cases decrease
significantly with growth in real
GNP. Therefore, for our period of examination, once we include all
DOJ and FTC antitrust
cases, it is unclear which hypothesis presented by Posner (1970)
will hold. During the Great
Depression, we expect Posner’s hypothesis (i), above, to hold;
however, this may be due
to industry codes formed before the NIRA and during the NIRA and
not necessarily the
contraction itself. In other words, a decrease in cases may be due
to the possible change
in regulators’ budgets and other contractionary effects, or it may
be due to the relaxation
of competition law. However, prior to the Great Depression, when
industry codes were in
effect, though collusion through such codes was not formally
government-supported, the
correlation between cases and GNP is unclear. If we are to argue
that industry codes that
allowed collusion, whether formally or not, led to a suspension of
antitrust enforcement, we
would expect antitrust cases to decline during the period the FTC
supports and sponsored
trade associations and the formation of industry codes.
10
Posner (1970) and Gallo, Dau-Schmidt, Craycraft, and Parker (2000)
analyze another
metric for enforcement: a civil or criminal remedy. While both
remedies involve sanction,
clearly a criminal remedy is of higher magnitude and higher
economic costs. Several have
investigated the deterrent effects of a civil versus criminal
remedy. Calkins (1997) finds
that civil fines have a smaller deterrent effect than a criminal
fine. Therefore, one might
hypothesize that a criminal remedy as opposed to civil remedy
indicates greater antitrust
enforcement. On a first look, Gallo, Dau-Schmidt, Craycraft, and
Parker (2000) find that
from 1925-1934, the DOJ pursued twice as many civil cases as
criminal cases. They also find
that of those imprisoned for antitrust violations, none were
imprisoned not only because of
price-fixing charges, but in addition to findings of intimidation,
threats or violence. From
1935-1939, they find that for 57 DOJ antitrust cases, 53 per cent
were civil. Therefore,
criminal cases clearly are not uncommon. We want to examine civil
versus criminal cases in
more detail, as the remedy and the type of case can reveal stricter
enforcement of antitrust
policy.
4 Data
We collect data on DOJ antitrust cases from the Annual Report of
the Attorney General
(Department of Justice (1920-1940)) and FTC antitrust cases from
the Annual Report of
the Federal Trade Commission (Federal Trade Commission
(1920-1944a). These Reports
provide very brief summaries of cases, providing at least one
defendant name, as well as
indications of the violation, good or service, industry, and
outcome. Additional DOJ case
information and the industries which commit violations is provided
in a book summarizing
federal antitrust cases Department of Justice (1936); however, not
every case initialized in
any given year is summarized in these publications. More detailed
explanations of DOJ cases
are provided in the CCH Trade Regulation Reporter (also known as
the CCH Bluebook)
(Commerce Clearing House (1932-1939)) and FTC cases in the Federal
Trade Commission
Decisions (Federal Trade Commission (1920-1944b)). These are the
sources used by Posner
11
(1970), which includes only cases finalized in a given year. Though
the data analysis becomes
more complicated when we include all cases instigated in a
particular year, whether finalized
or continued, we believe the fact a case was instituted and
information about the industry
and good for which the case belongs in that given period is
revealing information about
antitrust enforcement. Therefore, we include these additional cases
in our analysis, but
must be careful in how we use this information and account for
replication when relevant.
We complement our data collection from the Attorney General’s
Reports and the FTC
Reports with these three additional sources of data.
We consider the years 1925-1939 because we want to investigate
other non-legislative
antitrust policies and how the also affect antitrust enforcement
during the Great Depression.
We also include these years to compare enforcement outside of the
NIRA. Furthermore, the
fact that prior to the NIRA, the FTC encouraged trade associations
and sponsored Trade
Practice Conferences for industries to form codes, indicates that
the government, to some
degree, was exempting aspects of competition law prior to the NIRA.
Therefore, to some
extent, collusion was permitted in certain industries well before
the contraction. We want
to understand how the formation of industry codes before the Great
Depression affects
antitrust enforcement.
Finally, we include the years 1925-1939 for comparison to Posner
(1970)’s reporting of
data in five-year increments. Table 2 is a replication of parts of
tables from Posner (1970) of
DOJ and FTC total cases for the years 1925-1939. Table 3 lists the
total number of cases we
find described in the Attorney General Reports. Quick observation
reveals Posner’s totals
and the totals from these Attorney General Reports do not equate.
Posner considers only
cases finalized in a given year.
We keep record of all cases initialized in a given year, but
ultimately have to avoid
duplication for the continuance of the cases. The cases listed in
Table 3 include all cases
pending and initialized in a given year. Therefore, all cases
pending in the next year are
duplicated in the total for the following year. Also included in
Table 3 is the Attorney
General report of cases finalized in each year. These totals also
do not equate to Posner’s
12
Table 2: Total DOJ and FTC cases reported in Posner (1970) and to
be compared; Robinson-Patman Cases and FTC Trade Practice
Conferences to be determined.
Year DOJ Cases
1936 5 33 75 (1936-39) tbd
1937 7 18 tbd tbd
1938 10 28 tbd tbd
1939 31 31 tbd tbd
totals. We have yet to resolve the discrepancies. Based on Posner’s
totals, we expect a total
of 146 DOJ cases and 237 FTC cases to analyze over the period
1925-1939.
Table 4 lists the reported number of inquiries and formal
complaints by the FTC. Com-
paring the formal complaints to Posner’s total FTC cases shows an
obvious discrepancy for
the only year entered, 1930. Of the formal complaints reported by
the FTC in each year,
various violations are included. However, Posner only reports
restraints of trade cases.
Tables 5 and 6 provide appropriations to the Antitrust Division and
the FTC for each
year, with an itemization of salaries and general work for the FTC.
Also included for the
DOJ (currently unknown for the FTC) is additional remedy
information for DOJ cases,
including both civil and criminal fines, criminal convictions and
total number of months of
incarceration.
Potential problems with our analysis include dealing with the
following. First, as high-
lighted in Gallo, Dau-Schmidt, Craycraft, and Parker (2000), many
cases could result from
one initiated case that provided leads to other violations within
an industry. Therefore,
13
Table 3: DOJ Antitrust Division total civil and criminal cases
(Source: AG Reports) Year Total
(pending and init)
Civil finally det.
Crim finally det.
1925 59 29 30 11 8 3
1926 67 32 35 30 18 12
1927 45 29 16 23 15 8
1928 48 26 22 22 12 10
1929 43 26 17 13 8 5
1930 44 32 12 9 6 3
1931 39 29 10 21 14 6
1932 22 18 4 4 2 2
1933 25 21 4 6 4 2
1934 22 tbd tbd 8 tbd tbd
1935 tbd tbd tbd tbd tbd tbd
1936 31 18 13 10 7 3
1937 31 15 16 3 tbd tbd
1938 tbd tbd tbd tbd tbd tbd
1939 tbd tbd tbd tbd tbd tbd
Table 4: FTC total inquiries and formal complaints (Source: FTC
Reports) Year Total inquiries Formal Complaints
1925 tbd tbd
1926 tbd tbd
1927 tbd tbd
1928 tbd tbd
1929 tbd tbd
1930 1505 172
1931 tbd tbd
1932 tbd tbd
1933 tbd tbd
1934 tbd tbd
1935 tbd tbd
1936 tbd tbd
1937 tbd tbd
1938 tbd tbd
1939 tbd tbd
Table 5: DOJ Antitrust Division appropriated budget, total fines,
total convictions and total jail time (Source: AG Reports)
Year Budget Civil Fines Crim Fines Convictions Jail time (mo)
1925 $203,930 $67.90 $235,500 2 tbd
1926 $228,000 $297.58 $292,301 4 0
1927 $200,000 tbd $335,850 4 72
1928 $198,000 $1,725 $114,500 5 0
1929 tbd $3,658 $37,000 4 36
1930 $203,600 tbd $46,025 69 240.5
1931 tbd tbd tbd tbd tbd
1932 $204,160 tbd $100,865 100 tbd
1933 $150,000 tbd $222.10 2 tbd
1934 $153,000 tbd $559 15 48
1935 tbd tbd tbd tbd tbd
1936 tbd tbd tbd tbd tbd
1937 tbd tbd tbd tbd tbd
1938 tbd tbd tbd tbd tbd
1939 tbd tbd tbd tbd tbd
Table 6: FTC appropriated budget itemized by salaries and general
work (Source: FTC Reports)
Year Budget Salaries for Commissioners General Work
1925 tbd tbd tbd
1926 tbd tbd tbd
1927 tbd tbd tbd
1928 tbd tbd tbd
1929 tbd tbd tbd
1930 $1,495,821.69 $50,000 $1,390,971.82
1931 tbd tbd tbd
1932 tbd tbd tbd
1933 tbd tbd tbd
1934 tbd tbd tbd
1935 tbd tbd tbd
1936 tbd tbd tbd
1937 tbd tbd tbd
1938 tbd tbd tbd
1939 tbd tbd tbd
Date year, month and day case initiated and then decided
Length number of months from initiation to final decision
(including ap- peals)
Defendant names of all defendants, when known
Defendant type dummy variable whether defendant(s) was firm or
particular indi- vidual in that firm. If individual, what type
(director, owner, pres- ident, vice president, secretary/treasurer,
corporate officer, other)
Plaintiff name Typically US, but in cases where a private suit was
appealed, it may be a previous defendant
Industry in detail and by SIC code, broken up by division, major
group, industry, and four-digit SIC code
Alleged violation in detail and coded under 27 possible
violations
Court court where final decision made
Disposition Defendant’s plea and the court’s decision
Winner dummy variable = 1 if US; 0 otherwise.
Sanction explanation of the sanction in detail
Remedy dummy variable for civil (injunction, fine or both) or
criminal (incarceration, fine or both)
Fine amount of fine, if applicable (civil or criminal cases)
Jail time months of incarceration, if applicable (criminal cases
only)
DOJ budget appropriated budget for the Antitrust Division
FTC budget appropriated budget for the FTC (can be broken down into
salaries and general work)
lack of violations in other industries may be just due to lack of
information or tradeoff
in investigating cases and leads in another industry. Also,
controlling for the number of
employees is important since fewer employees clearly means fewer
cases can be prosecuted,
but such data currently is lacking; we will use budgets as a proxy
for the time being.
4.1 Variables
Table 7 summarizes the variables collected in reading the DOJ and
FTC court cases.
We use this data as metrics for investigating the degree of
antitrust enforcement in the years
just prior to and during the Great Depression. Next we present
hypotheses with respect to
each metric examined, and when possible, offer summaries of
findings by Posner.
16
4.1.1 Length of each case
The length of each case is determined by the number of months
between the initiation
and decision or executed remedy date of each case. We expect that
the longer the cases,
the stricter the antitrust enforcement. However, many cases end
with a consent decree
(agreement in civil cases) or nolo contendere (no contest plea in
criminal cases). In such
cases, the government wins, but the case length likely is very
short, since both entail out of
court settlements and thus no litigation. Therefore, the shorter
the case length in which the
government loses indicates more lax enforcement. However, we
examine how many cases
were settled and litigated. Posner (1970) finds that for the 59 DOJ
cases between 1925
and 1929, 44 were settled within six months and 15 were litigated
(25 per cent). For the
30 DOJ cases between 1930 and 1934, 20 were settled in six months
and 10 were litigated
(33 per cent). Finally, for the 57 DOJ cases between 1935 and 1939,
21 were settled within
six months and 27 were litigated (47 per cent). The percentage of
cases litigated appears
to increase, and pre-NIRA litigation appears lower than post-NIRA
litigation, offering no
evidence for more lax enforcement in terms of efforts to try a
case. With further analysis,
we can examine length for civil versus criminal cases and by
industry to determine which
types of violations and cases required or expended more effort of
the DOJ.
Table 8, replicated from Posner (1970), provides a breakdown in
years of five of DOJ
cases by civil and criminal, indicated the number of consent
judgments, and provides the
number of cases over lengths of time. We see that in early years,
more cases were settled
within six months and more consent decrees were issued, indicating
settlement agreements
with the DOJ. During the Hoover administration of tough enforcement
and the beginning
for most of the NIRA, at least four cases last over two years.
While Posner’s examination
is useful, a more detailed analysis of length by year and
interacted with other variables will
reveal more information about enforcement during this period.
17
Table 8: DOJ cases: length and number (replicated from Posner
(1970)) DOJ variables 1925-1929 1930-1934 1935-1939
Civil cases (in mo.) 31 40 62 Criminal cases (in mo.) 32 51
26
No. 12 mo. or less (civil)
No. 13-24 mo. (civil) 1 3
No. 25-36 mo. (civil) 3 2
No. 36-48 mo. (civil) 4 1
No. 49-60 mo. (civil) 1
No. 61-72 mo. (civil) 1 1 1
No. 73-84 mo. (civil)
No. 125 mo. or more (civil) 4
No. cases settled w/in 6 mo. 44 20 21 No. Consent judgments (civil)
33 12 17
4.1.2 Defendant type
While we often do not have details of every defendant, the number
of and types of
defendants offer insight into the strictness of the DOJ and FTC in
which parties they deem
violators. Evidence of individuals, rather than just the firm,
being prosecuted might indicate
stricter antitrust enforcement. A more intense antitrust enforce
would seem to go after not
just a firm, but individuals within a firm, as well, if such
violation warrants individual
sanctions. Therefore, like Gallo, Dau-Schmidt, Craycraft, and
Parker (2000) we determine
what type of entity the DOJ and FTC pursue. Becker and other
theories of punishment
argues that firms are the more efficient violator to pursue in
terms of deterrence and cost
because of the judgment proof problem with individuals and the
greater likelihood a firm can
pay a fine, so that incentives of the firm are in fact affected, as
well as costs of imprisonment
are avoided. Others support remedies involving specific individuals
in a firm to increase
incentives by firm employees to adhere to antitrust policies.
18
4.1.3 Industry
We break down the good or service for each case by the Standard
Industrial Classifica-
tion (SIC) four-digit codes, as well as the division. Ten divisions
(A-J) make up the SIC
groupings, then the hierarchy of each four-digit code is broken
down in major groups (first
two digits), industry group (third digit) and finally, the good or
service itself is indicated by
the fourth digit. We break down each good and service to compare
industries and general
types of goods/services with those for which participated in the
NIRA.
Clearly many large industries were favored by the NIRA, and we want
to understand
if these industries also were favored in non-NIRA allowed
anticompetitive behavior, re-
flected by fewer antitrust cases in these industries, or whether
regulators focused less on
the industry and more an specific violations not allowed under the
NIRA. Additionally, as
mentioned above, evidence suggests some industries continued to
follow their industry codes
post-NIRA, in which case such behavior would violate antitrust
laws. If we see few antitrust
cases in industries known to continue following industry codes of
collusive behavior (and
the specific behaviors in those codes not being prosecuted), then
we have some evidence of
lax antitrust enforcement.
4.1.4 Alleged violation
We categorize six types of antitrust violations, one of which is a
violation for threat-
ening behavior (which can lead to criminal prosecution). Between
these seven types
of violations, we have 29 total possible violations (assigned a
value 0-28). These 29
violations were categorized in part by the violations considered in
Posner (1970) and
Gallo, Dau-Schmidt, Craycraft, and Parker (2000). We will
investigate each violation as
well as determine the frequency of each type of violation to
determine the focus of antitrust
authorities. We also will combine this information with the NIRA
codes to determine if we
find a decline in cases for specific violations and if that decline
correlates with industries
and the particular codes of collusion allowed by the government,
whether informally prior to
the NIRA and formally during the two years of the NIRA. As
indicated above, post-NIRA,
19
under a hypothesis of strict antitrust enforcement, we expect an
increase in prosecutions
for violations that were allowed by the NIRA, particularly in those
industries that openly
continue to adhere to their codes. Table 9 outlines the 29
violations.
4.1.5 Court and/or Location
By examining location and the court, we hope to understand if
specific courts favored
one side or were more lenient (or harsher) for certain alleged
violations.
4.1.6 Disposition, Winner, Sanction and Remedy
With respect to the type of pleas, we might see a decrease in the
number of guilty pleas
if the penalties are higher (stricter enforcement) because
defendants might want to avoid
harsh penalties. This is a theory presented by Snyder (1990) (page
439).
We also breakdown cases by civil and criminal. We expect criminal
violations to be
extreme and hope to get a clear picture of what other antitrust
violations are part of the
criminal cases, as criminal sanctions are stricter than civil
sanction and can be an indicator
or tougher enforcement. During this period, the maximum statutory
penalty for antitrust
violations under the Sherman Act was $5, 000 for firms and
individuals, with a maximum
one year incarceration (misdemeanor) for individuals. We
hypothesize that the closer to
these limits the regulator got, the stricter the antitrust
enforcement.
We would like to investigate aspects of the data such as (i)
violations and industries
enforced by regulators both during and outside the NIRA; (ii)
violations and industries
enforced by the courts (wins/losses by violation) both during and
outside the NIRA; (iii)
which violations were tried under criminal statutes; (iv) trends in
remedies, fines and in-
carceration over time and by industry and violation; (v) types of
defendants by violation
and industry; (vi) length of cases both during and outside the
NIRA; and (vii) if specific
courts favor regulators.
Category Violation
Horizontal (per se) restraints
1 - price fixing or agreement to maintain fixed min. prices 2 -
colluding on something other than price 3 - fixed output levels
(production or sales quota) 4 - trade association or equivalent 5 -
policing, fines, audits 6 - territory allocation schemes 7 -
customer allocation schemes 8 - bid-rigging with government as a
buyer 9 - other bid-rigging
Monopolization
10 - single-firm monopolies 11 - single-firm patents 12 -
multiple-firm monopolies 13 - multiple firm patents 14 - cases with
dissolution or significant divestiture
Exclusionary practices
15 - price discrimination 16 - boycott 17 - reciprocity 18 - tying
arrangements 19 - misuse of patents or threatening patent action 20
- sabotaging competitors 21 - exclusive dealing
Vertical restraints 22 - single-firm RPM 23 - multiple-firm RPM 24
- restrictive requirements on franchisers or dealers - something
other than RPM
25 - Horizontal merger violations
21
5 Proposed analysis
Beyond the basic statistical evaluation of antitrust cases between
1925-1939, we can
also consider what explains whether the United States (DOJ or FTC)
wins an antitrust
case, looking at factors such as the industry, the violation, the
Court, change in budget.
Using an event study, we also can examine the effect of the change
in antitrust policy to
determine if the number of cases statistically change during the
NIRA compared to nonevent
periods before and after the NIRA. Finally, we hope to combine our
case data with industry
code data to determine how the DOJ and FTC changed antitrust
enforcement in response
to the NIRA codes for specific industries and for specific
(anticompetitive) behavior. For
example, did regulators shift away to industries that not sign
industry codes, or did they
merely focus more on specific violations not covered by industry
codes, but favored no
particular industry? We can break down all of these analyses by
civil and criminal cases,
or include such as an explanatory variable.
6 Conclusion
We offer an empirical study of antitrust enforcement by the
Department of Justice
(DOJ) and Federal Trade Commission (FTC) during the New Deal. The
goal of this
research is to address arguments that antitrust policies during the
Great Depression were
harmfully lax or even “suspended”. These arguments stem from the
National Industrial
Recovery Act (NIRA), lasting from 1933 to 1935, that allowed
industries to collude in
specific ways under specific rules. Therefore, on the surface, the
NIRA indicates antitrust
policies were nonexistent during this period. Other non-legislative
antitrust actions by
regulators and the courts also affected antitrust policy. However,
changes in antitrust policy
do not imply no enforcement of competition law during this period.
We seek to resolve this
growing assumption in the literature by creating a detailed
dataset, and then carefully
examining, federal antitrust cases from 1925-1939. Resolving this
growing assumption of
nonexistent antitrust enforcement is important because some base
conclusions on the extent
22
of the Great Depression and output on an assumption of suspended
antitrust policy. Such
policy implications during this period have been paralleled to
potential policy changes in
today’s economic environment. Our analysis provides a detailed look
at the industries
and alleged violations investigated and prosecuted by the DOJ and
FTC, as well as other
non-legislative antitrust policies in place, the size of the
regulatory agencies, the resulting
remedies, and length of cases. We examine the years 1925-1939 due
to vacillating antitrust
enforcement perspectives of the government between these years, as
well as to compare
to periods outside of the NIRA. Complementing our case data with
specifics of the NIRA
Codes of Fair Competition (rules of collusion for each industry),
as well as data on other non-
legislative antitrust policy, we determine factors that explain
antitrust enforcement, thereby
providing a clearer picture of exactly how antitrust enforcement
responded to changes in
antitrust policy during this fragile period. We still are entering
data and further analysis
will continue in the ways indicated above.
23
References
Alexander, B. J. (1997): “Failed Cooperation in Heterogeneous
Industries Under the National Recovery Administration,” The Journal
of Economic History, 57(2), 322–344.
Alexander, B. J., and G. D. Libecab (2000): “The Effect of Cost
Heterogeneity in the Success and Failure of the New Deal’s
Agricultural and Industrial Programs,” Explo- rations in Economic
History, 37, 370–400.
Baker, J. B. (2003): “The Case for Antitrust Enforcement,” The
Journal of Economic Perspectives, 17(4), 27–50.
Berle, A., and G. C. Means (1932): The Modern Corporation and
Private Property. New York: Macmillan.
Bittlingmayer, G. (1995): “Output and Stock Prices When Antitrust
Is Suspended: The Effects of the NIRA ,” in “The Causes and
Consequences of Antitrust: A Public Choice Perspective”, ed. by F.
S. McChesney, and W. S. I. (eds.)., chap. 17, pp. 297–318.
University of Chicago Press.
Calkins, S. (1997): “Corporate Compliance and the Antitrust
Agencies’ Bi-Modal Penal- ties,” Law and Contemporary Problems, 60,
127–167.
Cole, H. L., and L. E. Ohanian (2004): “New Deal Policies and the
Persistence of the Great Depression: A General Equilibrium
Analysis,” Journal of Political Economy, 112(4), 779–816.
Commerce Clearing House (1932-1939): Trade Regulation Reporter,
United States An- titrust Case Summaries.
Cox, C. C. (1981): “Monopoly Explanations of the Great Depression
and Public Policies Toward Business,” in The Great Depression
Revisisted, ed. by K. B. (ed), chap. 8, pp. 174–207.
Springer.
Dearing, C. L., P. T. Homan, L. L. Lorwin, and L. S. Lyon (1934):
The ABC of the NRA. Washington, D.C., Brookings Institution.
Department of Justice (1920-1940): Annual Report of the Attorney
General of the United States for the Fiscal Year XXXX.
(1936): The Federal Antitrust Laws With Summary Of Cases Instituted
By The United States And Lists Of Cases Decided Thereunder ......
Nabu Press.
Eggertsson, G. B. (2012): “Was the New Deal Contractionary?,”
American Economic Review, 102, 524–555.
Federal Trade Commission (1920-1944a): Annual Report of the Federal
Trade Com- mission for the Fiscal Year Ended June 30, XXXX.
(1920-1944b): Federal Trade Commission Decisions.
24
Gallo, J. C., K. Dau-Schmidt, J. L. Craycraft, and C. J. Parker
(2000): “De- partment of Justic Antitrust Enforcement 1955-1997: An
Empirical Study,” Review of Industrial Organization, 17,
75–133.
Hannsgen, G., and D. B. Papadimitriou (2009): “Lessons from the New
Deal: Did the New Deal Prolong or Worsen the Great Depression?,”
Working Paper No. 581.
Hawley, E. W. (1966): The New Deal and the Problem of Monopoly: A
Study in Economic Ambivalence. Fordham University Press.
Nathan, O. (1935): “The N.I.R.A. and Stabilization,” The American
Economic Review, 25(1), 44–58.
Peltzman, S. (1976): “Toward a General Theory of Regulation,”
Journal of Law and Economics, 19, 211–240.
Poole, W. (1981): “Comments on “Monopoly Explanations of the Great
Depression and Public Policies Toward Business”,” in The Great
Depression Revisisted, ed. by K. B. (ed), chap. 9, pp. 208–219.
Springer.
Posner, R. A. (1970): “A Statistical Study of Antitrust
Enforcement,” Journal of Law and Economics, 13(2), 365–419.
Taylor, J. E. (2007): “Cartel Code Attributes and Cartel
Perforance: An Industry-Level Analysis of the National Industrial
Recovery Act,” Journal of Law and Economics, 50, 597–623.
25
Introduction
Related Literature
Proposed analysis