Tariff politics and congressional elections: exploring the Cannon Thesis
| Date | 01 July 2017 |
| Author | Kenneth S Lowande,Jeffery A Jenkins,Andrew J Clarke |
| DOI | 10.1177/0951629816647801 |
| Published date | 01 July 2017 |
| Subject Matter | Articles |
Article
Journal of Theoretical Politics
2017, Vol.29(3) 382–414
ÓThe Author(s) 2016
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0951629816647801
journals.sagepub.com/home/jtp
Tariff politics and
congressional elections:
exploring the Cannon Thesis
Andrew J Clarke
University of Virginia, USA
Jeffery A Jenkins
University of Virginia, USA
Kenneth S Lowande
WashingtonUniversity in St. Louis, USA
Abstract
While a number of studies have examined the politics of tariff decision-making in the United
States, little work has examined the subsequent political effects of tariff policy. We help fill this
gap in the literature by analyzing—both theoretically and empirically—the electoral implications
of tariff revision. Specifically, we investigate the veracity of the Cannon Thesis—the proposition
advanced by Speaker Joe Cannon in 1910 that the majority party in the U.S. House was punished
when it made major revisions to the tariff. We find that from 1877 to 1934 major tariff revisions
were, on average, associated with a significant loss of votes for majority-party members—both
regionally and nationally—that translated into a loss of House seats. We find support for the
notion that major tariff revisions generated inordinate uncertainty among various business inter-
ests, which the opposition party could then use (by leveraging fear and market instability) to
mobilize its base and gain ground in the following election. Our results provide a new explanation
for the delegation of tariff policymaking to the executive branch.
Keywords
Congressional elections; rollcall votes; tariffs
Corresponding author:
Jeffery A Jenkins, Universityof Virginia, 1540 Jefferson Park Avenue, Charlottesville,VA 22903, USA.
Email: jajenkins@virginia.edu
1. Introduction
The tariff—and international trade more generally—has been among the most con-
tentious issues in American politics since the nation’s inception. This is due, in
part, to the scope of international trade policymaking, as tariff schedules and trade
restrictions involve—whether directly or indirectly—nearly every economic sector
in the United States. As a result, various agricultural, industrial, and manufactur-
ing interests have taken a keen interest in trade politics and the construction (and
proposed revision) of the tariff.
The centrality of the tariff as a national political-economic issue was paramount
in the pre-World War II era. From the early 19th century (beginning with the
Tariff of 1816, the nation’s first protective tariff) through the beginning of the New
Deal era (and the Reciprocal Trade Agreement Act of 1934), tariff politics pitted
various sectional interests against one other, divided the parties (Democrats from
Whigs/Republicans), and often featured heavily in national election campaigns
(especially after Reconstruction).
1
During this time, the locus of tariff policymak-
ing was in Congress, as members spent considerable time debating the tenets of free
trade versus protectionism and responding to district/state interests that sought to
entrench the status quo or revise the overall tariff schedule (Democrats typically
downward, Republicans typically upward). By 1934, the complexity of tariff pol-
icymaking, and escalating costs therein, pushed members of Congress to delegate
international trade issues to the Executive Branch.
2
From an academic perspective, the tariff has been the focusof numerous studies.
Indeed, as Goldstein (1993: 4) notes: ‘Few political phenomena have been studied
as thoroughly as tariff policy.’ The range of scholarship has encompassed tradi-
tional histories (Stanwood, 1903; Taussig, 1931), analytical histories (Bensel, 2000:
Chapter 7; Goldstein, 1993; Lake, 1988), ‘publicchoice’ analyses (Conybeare, 1991;
Pincus, 1977), and more ‘normal science’ (formally or non-formally inspired) quan-
titative approaches (Brady et al., 2002; Epstein and O’Halloran, 1996; Fordham
and McKeown, 2003; Hansen, 1990; Hiscox, 2002; Weller, 2009). And while vari-
ous studies have examined the politics of tariff decision-making—often as they
related to the construction or revision of particular tariffs
3
—and how (a) sectional
interests and congressional parties lined up, and (b) individual members voted, little
work has examined systematically the subsequent political effects of tariff policy.
Stated differently, tariff policymaking—at either the aggregate or individual level—
is often considered the phenomenon to be explained rather than a factor explaining
some other phenomenon.
This study explicitly follows the latter path, by considering the tariff as a poten-
tial determinant in national election outcomes. Specifically, we explore whether a
major tariff revision during the period spanning 1815–1934 affected the subsequent
electoral fortunes of the majority party in the House of Representatives. Our inter-
est in this question is motivated by statements made by House Speaker Joe
Cannon (R-IL) in 1910, days before the November midterm elections (which
would result in his Republicans being swept from power). Speaking to a group of
Clarke et al. 383
reporters, Cannon voiced a belief about the causal relationship between major tar-
iff legislation and the electoral performance of the House majority; specifically,
that the enactment of a major new tariff prior to the November elections resulted
in the majority party ‘getting licked’ (Dallas Morning News, 11/8/1910, 1; see also
Washington Post, 11/7/1910, 3).
4
He then provided a series of examples going back
nearly 30 years to support his proposition. Thus, in Cannon’s view, tariff reform
constituted the ‘third rail’ of electoral politics for the majority party in the House
at that time.
Cannon’s proposition—or ‘thesis,’ as we will call it
5
—has never been explored
in depth or tested systematically. Some scholars have presented anecdotal accounts
to suggest that a particular tariff led to the enacting coalition’s (majority party’s)
defeat in the ensuing elections—the McKinley Tariff of 1890, which resulted in the
Republicans suffering significant losses and yielding majority control of the House
to the Democrats, is an oft-told story (Stanwood, 1903; Taussig, 1931). But, by and
large, scholars of modern social science have ignored the connection between tariffs
and elections almost entirely.
6
We seek to fill this gap in the literature by engaging
the Cannon Thesis directly with some straightforward time-series analyses.
The paper proceeds as follows. In the next section, we explore the Cannon Thesis
in detail, by outlining Cannon’s logic as to why changes in the tariff (regardless of
whether it was raised or lowered) should have produced electoral fallout for the
majority party. Given this logic, we present a model to identify conditions under
which a ‘rational’ majority party might revise the tariff—knowing full well the elec-
toral consequences. In the third section, we identify major tariff revisions between
1815 and 1934, and specify a series of time-series models by electoral cycle (i.e., elec-
tions to a given Congress) to examine the impact that they had on majority-party
vote totals, controlling for other important factors/covariates; we find that for the
1877–1934 period, major tariff revisions were associated with a significant loss of
votes—both regionally and nationally—for the majority party, which translated
into a loss of House seats. In the fourth section, we show (a) that these electoral
consequences do not merely coincide with the cyclical effects of midterm elections
and (b) failed attempts at tariff reform do not generate electoral consequences of
equal magnitude. In the fifth section, we explore individual-level electoral conse-
quences for a given case, the Wilson–Gorman Tariff of 1894, and find that a vote
for the tariff increased a House member’s probability of defeat. In the final section,
we conclude by discussing the implications of our work for further study.
2. The Cannon Thesis in historical and theoretical context
Joseph Gurney Cannon was a colorful character who epitomized the classic Gilded
Age politician.
7
A representative from downstate Illinois, Cannon was a cigar-
smoking, hard-drinking ‘everyman’ who eschewed progressive reform and rose to
the top of an institution that was built on machine-style politics and partisan
patronage. First elected in 1872, ‘Uncle Joe’ became Speaker of the House in 1903,
384 Journal of Theoretical Politics 29(3)
To continue reading
Request your trial