[ Pobierz całość w formacie PDF ]
powers (SOP) model. Admittedly, this is not the only theory of influence on
agency policy. Another theory showing how agencies respond to competing
principals is that of common agency (Dixit 1995; Dixit et al. 1997). Specifi-
cally, it is a theory of a multilateral relationship in which several principals
simultaneously try to influence the actions of an agent (Dixit et al. 1997).
For our purposes, this formulation better describes the process herein
of multiprincipals with common goals acting sequentially. The IRS regional
directors are appointed by the IRS commissioner, an appointee of the president
subject to confi rmation by the Senate. I assume that the regional directors
will carry out the policy preferences of the president because the president
has the power to fi re and replace a wayward director.
The regional directors should use their ability to shift audits between
states and districts to most effectively carry out the president s policy, subject
to the constraint of the legislative branch. As Figure 5.1 (next page) shows,
the SOP model suggests that a core of legislative and executive branch
preferences constrain agency policy choices (Hammond and Knott 1996;
Hammond and Miller 1987). The figure implies that in a liberal-conservative
dimension, agencies may propose any policy within a legislative-executive
core because that policy will not be overturned through statutory means
(Hammond and Knott 1996). If the president, P, is relatively extreme, one
boundary of the legislative-executive core is defi ned by the median House
member, H , or the median senator, S , whoever is furthest from the presi-
m m
dent. The crucial veto-override legislator in the House, H , or Senate, S ,
vo vo
whoever is closer to the president, defi nes the other boundary of the core.
Boundaries defi ned by this core represent the preferences of the dominant
national coalition.
78 GETTING A POOR RETURN
Legislative-Executive Core
Liberal Conservative
P H S H S
vo vo m m
Executive Appointee Target
P = President
H = 290th Most Liberal House Member (key veto-override representative)
vo
S = 67th Most Liberal Senator (key veto-override senator)
vo
H = Median Representative
m
S = Median Senator
m
Figure 5.1 Executive-Appointee Target Veto-Override Point
An agency, or regional director, that establishes a policy outside the
core will be overturned through a statutory revision at some point within the
core. If the SOP model is accurate, executive agencies will establish policies
at the boundary of the legislative-executive core closest to the president,
shifting as it shifts (Shipan 1998).
Based on this core dynamic, each regional director ought to establish
policy at the exact same point. However, there is still a need to control for
nonpolitical factors such as regional wealth disparities, or even regional or
local political factors. For example, IRS policy can directly impact a state s
economy and local politicians would ignore the effect of IRS audits at their
peril. Therefore, other political actors, such as governors, may play a role
in the calculations of regional directors.
Courts and the Legal Set within the Separation of Powers Framework
Courts should also induce variation in agency policy through a legal set
that may or may not overlap with the legislative-executive core. If the legal
INFLUENCES ON THE IRS 79
set does not overlap with the core, then judicial review presents an addi-
tional constraint on agency choice. A regional director may not establish a
policy at the boundary of the core because the director risks a court-ordered
policy anywhere within the legal set. Instead, the regional director should
establish a policy at the boundary of the legal set closest to the president,
if it lies within the core. Such a policy will survive any legislative efforts to
overturn it, and will also pass judicial review. It represents the best policy
the president can obtain. For a regional director of the IRS, there are two
primary trial court legal sets as demonstrated in Figure 5.2. Here TC repre-
sents the ideal policy of the Tax Court, and DC represents the ideal policy
of any particular federal district court.
Of course, how does the agency determine the policy preferences or
legal set for a trail court? Several court scholars examining appellate court
decision making have established the median justice as pivotal in calculating
court preferences (Hausegger and Haynie 2003; Martin, Quinn, and Epstein
2004). The median justice is crucial on multimember appellate courts where
every justice has a vote, with a majority decision determining the outcome.
[ Pobierz całość w formacie PDF ]