Tax subsidies vs. non-tax subsidies

[Volokh, Eugene]

Alan Brownstein writes:

> First, would economists distinguish between the government giving a 
> religious institution cash subsidies (enough to hire a person for a 
> particular job with appropriate benefits) to hire someone to 
> provide some secular assistance to a religious school or charity and
the 
> government hiring a person that satisfied the religious institution's
employment 
> criteria (including religious belief) and assigning the employee to
the 
> institution to provide the same assistance? Assuming that 
> this is a general program available to both secular and religious non 
> governmental providers, if the former is constitutional, is the latter
constitutional?

	It seems to me that the Establishment Clause embodies several
different principles -- for instance, (1) no religious decisions by the
government (cf. Milivojevich), (2) no endorsement of particular
religions, (3) no religious discrimination by the government, (4) no
improper benefits flowing to religious institutions (cf. Mitchell v.
Helms), and so on.  But if we're working with principle 4, and the
concern is the flow of money or of other valuable things, then we
shouldn't treat economically identical things differently for those
purposes.  And indeed providing in-kind services should be the same as
providing money or a tax exemption.  For instance, if the government
decides to station police officers in all schools to combat school
violence -- officers who will enforce secular law, not the school's
religious rules -- then it should be free to station such officers in
religious schools as well as secular schools, just as it should be free
to let religious schools participate in voucher programs, in the
charitable tax exemption, and so on.  My view is that this "no money
flow" principle shouldn't have indeped

	But an action may be just fine under principle 4 (either because
it only supports the religious organization's secular activities, see
the Mitchell plurality and concurrence, or because it involves a private
choice program, see Zelman, or because it fits within Rosenberger) but
still violate another principle, such as 3.  The problem with the
government's hiring a person based on the religious institution's
employment criteria, if there is a problem with it (and I think there
is), is that this involves the government itself discriminating based on
religion.  But, as the stationing-the-police-officer-at-all-schools
example shows, the objection is precisely the discrimination -- not the
improper flow of benefits.

> Second, the answer to these exemption/subsidy issues may not be either
of 
> the two alternatives that you suggest (as Marty says, notwithstanding
what 
> economists think). Working out a framework to reconcile conflicting
values 
> in the religion clause area may involve line drawing that is in part 
> arbitrary and in part represents an assessment of risks to be avoided
and 
> goals to be achieved. Tax exemptions as Marty argues may be less prone
to 
> manipulation and government favoritism, taxing religious institutions
may 
> be viewed as particularly threatening to religious autonomy, tax
deductions 
> are more consistent with a tradition of religious voluntarism than
subsidies etc.

	Maybe -- but maybe not.  Given the strings that are regularly
attached to tax exemptions (see, e.g., the 501(c)(3) rules that bar
recipients from engaging in political activity, the Bob Jones University
scenario, and so on), it seems to me that the case for distinguishing
tax exemptions from subsidies is fairly weak.  And this is especially so
when the consequences of such a distinction are mandatory discrimination
against religious institutions (even if the ostensible purpose for this
discrimination, maintaining the institutions' autonomy, is laudable).
> 
> Or to put the issue in a broader perspective, many religious
communities, 
> particularly minority faiths, need both independence and autonomy (for
want 
> of a better term call it reserved private space) and the ability to 
> participate with others in the larger community, economy, society, and

> polity (call that access to public space).  Serving both objectives 
> requires putting some limits on turning government resources 
> (and even some private resources e.g. corporate commercial employers)
into reserved 
> private space. Distinguishing between tax exemptions, vouchers, tax 
> deductions, direct subsidies etc are arguably imprecise but 
> useful ways to accomplish those objectives. Certainly one may argue
that 
> compared to many other societies, religious liberty and equality (and
religion 
> generally) have flourished in the U.S. during the period in which
those 
> distinctions have been recognized.

	Certainly one may argue that -- but it's hard to see much
evidence for a causal connection there.  Nor should one ignore, I think,
the burden on religious individuals caused by the exclusion of religious
institutions from generally available benefit programs.  When the
government selectively funds only secular alternatives (schools,
rehabilitation programs, etc.), it pressures people to use those
alternatives (which will be free, because government-subsidized) instead
of the now relatively more expensive religious alternatives.

	Schools provide the clearest example.  The Establishment Clause
case against school choice often relies on the argument that vouchers
will have strings, and religious schools will thus be pressured to
accept those strings..

	But the absence of vouchers causes its own sort of pressure:  A
family that has paid its taxes, including a large chunk that goes to
fund secular schools, will now be pressured to send their kids to
secular school -- even if that school provides a philosophy or an
environment that is hostile to the family's religious preferences --
instead of the religious school.

	Now I don't think that this pressure is so improper that it
should lead school choice to be a constitutional mandate.  But I do
think that before we conclude that school choice programs that
evenhandedly include religious schools -- and other evenhanded benefit
programs -- are unconstitutional because they can compromise the
religious institutions' autonomy, we should consider how they actually
*further* religious users' autonomy.

	Eugene



	I'm not in principle against the use of proxies for important
but hard to measure variables, if the variable really is hard to
measure, and the proxy really is good.  But I'm not sure that this is
really so.