Revisiting the Two Presidencies

• Author: Bryan W. Marshall,Richard L. Pacelle
• Published date: 01 January 2005
• Date: 01 January 2005
• DOI: http://doi.org/10.1177/1532673X04266816
• Subject Matter: Articles

10.1177/1532673X04266816
AMERICAN POLITICS RESEARCH / January 2005
Marshall, Pacelle / REVISITING THE TWO PRESIDENCIES
REVISITING THE TWO PRESIDENCIES
The Strategic Use of Executive Orders
BRYAN W. MARSHALL
Miami University
RICHARD L. PACELLE JR.
Georgia Southern University
Recent scholarly attention has considerably advanced our understanding of executive orders. We
argue that executive orders represent a valuable opportunity to assess the relevance of the two
presidencies and the distinction between foreign and domestic policies. The over-time analysis
(1953-1997)demonstrates significant differences in the effects for most of the variables explain-
ing executive orders depending on the issue area. For example, we find that the president’s share
of congressionalparty seats significantly affects executive orders on domestic policy, but no such
effects are found on foreign policy. This result, as well as many others in the analysis, illustrates
that factors shaping the executive’s ability to influence policy in Congress differ substantially in
their effects on domestic as compared to foreign policy executive orders. We infer from our anal-
ysis that the two-presidencies distinction remains a useful one, at least for understanding
executive orders.
Keywords: executive order; two presidencies; presidential power; congressional-presiden-
tial relations
Spanning several decades, scholars have been intrigued by questions
related to presidential power and the conditions that foster it (Herring,
1940; Neustadt, 1962; Rossiter, 1956; Wildavsky, 1969). However,
the task of measuring and explaining presidential power remains one
of the greatest challenges (Andres & Griffin, 2002; Lindsay & Steger,
1993). It is certainly true that scholars have significantly advanced our
understanding of the executives’ tools of influence in areas such as
public approval and the veto pen (Brace & Hinckley, 1992; Cameron,
Authors’Note: An earlier version of this research was presented at the annual American Political
Science Association Meetings, San Francisco, California, August 30-September 2, 2001. We wish
to thank Marc Hendershot, Andrea Pyatt, Jason Roberts, and Fenton Martin for their invaluable
assistance. We would also like to thank the anonymous reviewers for their suggestions.
AMERICAN POLITICS RESEARCH, Vol. 33 No. 1, January 2005 81-105
DOI: 10.1177/1532673X04266816
© 2005 Sage Publications
81

---

82
AMERICAN POLITICS RESEARCH / January 2005
2000; Canes-Wrone, 2001; Krehbiel, 1998). But compared to these
areas of presidential inquiry, executive orders represent a relatively
understudied source of behavior to assess theories of presidential
influence.
Much of the extant literature focuses on a few key considerations in
explaining executive orders. For example, some work suggests that
executive orders respond to presidency-centered conditions, whereas
other findings imply that presidents use executive orders strategically
in response to the favorable (or unfavorable) governing circumstances
associated with Congress (see Krause & Cohen, 2000; King & Ragsdale,
1988;andDeering&Maltzman,1999,respectively).Although our anal-
ysis is sensitive to the considerations in this literature, we move
beyond these explanations by focusing on how presidents issue execu-
tive orders across policy areas.
In particular, we employ presidential decision making on executive
orders to assess the two-presidencies thesis—one of the perennial
questions in the literature on presidential power. Whereas previous
explanations of the two presidencies focused almost exclusively on a
president’s roll call success with Congress, we argue that the two-
presidencies hypothesis should be tested more generally across a
range of relevant executive behaviors. The analysis offers some
unique findings demonstrating that executive orders are used differ-
ently across foreign and domestic policy. The model results on execu-
tive orders are not only consistent with the two-presidencies thesis but
also support our theoretical claims that greater congressional delega-
tion and executive discretion are likely in foreign policy as compared
to domestic policy.
PERSPECTIVES ON EXPLAINING EXECUTIVE ORDERS
Scholars have offered a number of explanations for presidential
decisions to issue executive orders (Deering & Maltzman, 1999;
Krause & Cohen, 1997; Krause & Cohen, 2000; Mayer, 1999;
Wigton, 1996). One central question addressed in this literature is
whether executive orders are used as a strategic policy tool designed
unilaterally to short-circuit the traditional legislative process. Deering
and Maltzman (1999, p. 779) find evidence supportive of this strategic

---

Marshall, Pacelle / REVISITING THE TWO PRESIDENCIES
83
view by showing that presidents use significantly more executive
orders the greater their ideological positions are from the House and
Senate medians. They also uncover evidence consistent with the idea
that under certain conditions executive orders serve to complement
presidential policy positions in Congress. In particular, Deering and
Maltzman show significant increases in executive orders when the
veto pivots in both houses make it less likely that the president’s policy
position can be overturned.
Other studies have made important contributions to our under-
standing of executive orders. Indeed, some research has argued that
executive orders respond primarily to intra-institutional conditions
such as the development of the plural presidency and other president-
centered factors (King & Ragsdale, 1988, p. 121; Krause & Cohen,
2000, pp. 89-94; Light, 1982). Krause and Cohen (1997, p. 470) also
infer from their findings that executive orders complement congres-
sional actions rather than serving as a strategic policy tool. More
recently, these scholars demonstrate a sharp change in the use of exec-
utive orders. Before 1969, president-centered factors were found to be
the most important in explaining executive orders, but after, the
impact of inter-institutional factors dominated (Krause & Cohen,
2000, pp. 109-111).
Mayer’s (1999, 2001) results explaining executive orders seem to
present a quandary for the strategic view. On one hand, Mayer’s find-
ings suggest that presidents issue more executive orders to circumvent
Congress when their approval ratings are low—presumably a less
favorable condition for legislating. But on the other hand, presidents
issue significantly more orders under unified government—presum-
ably a more favorable condition for legislating. Mayer (2001) con-
cludes, “If . . . presidents use executive orders to compensate for politi-
cal weakness in other areas, why does low popularity stimulate more
orders while weakness in Congress does not?” (p. 101). Similarly,
Shull’s (1997) over-time analysis shows that unified government
increases the frequency of executive orders, from which he infers that
orders are used to complement or reinforce legislative successes in
Congress. Together, these findings offer important exceptions to the
strategic view that suggests presidential decisions to issue orders are
designed solely to circumvent the legislative process.1

---

84
AMERICAN POLITICS RESEARCH / January 2005
Although this research has fueled an important debate toward
understanding presidential behavior, it misses a valuable opportunity
to probe into other relevant questions. Notably absent from these stud-
ies on executive orders has been the question of the two presidencies.
In our view, presidential decisions to issue executive orders represent
a valuable context to revisit Wildavsky’s (1969) two-presidencies the-
sis. Indeed, our analysis is designed to assess whether differences
exist in the issuance of executive orders across foreign and domestic
policies.
THE TWO PRESIDENCIES
Wildavsky’s (1969) classic study showed that the legislative suc-
cess rate of presidents in foreign and defense policy was nearly twice
as high as compared to domestic policy. The existence of this major
gap clearly supported his two-presidencies argument. Presidents
wielded much greater influence in foreign policy, whereas Congress
seemed more willing to assert its institutional powers in the domestic
arena. Subsequent analyses of the gap in presidential influence have
been operationalized exclusively in terms of roll call success
(Edwards, 1989). Yet we would argue that Wildavsky’s (1969) ideas
regarding the two presidencies were not just applicable to the narrow
interaction between Congress and the president on roll calls. Rather,
Wildavsky’s arguments explained presidential dominance in foreign
affairs more generally. Although roll call success is certainly one
important measure of influence, the use of executive orders represents
another.
Wildavsky (1969) contended that executive dominance in foreign
policy was rooted in the presidency’s institutional and informational
advantages relative to Congress as well as a general consensus regard-
ing the means to protect national interests in the dangerous Cold War
environment fol owing World War II. The substantive stakes and inter-
connectedness of foreign policy necessitated that presidents actively
engage in all facets of international affairs. Unlike domestic policy, a
single foreign policy failure could lead to an array of catastrophic
disasters.2 Wildavsky also suggested that policy preferences were
more varied in the domestic arena and therefore had a much greater

---

Marshall, Pacelle / REVISITING THE TWO...