Ayn Rand is noted for her outspoken political philosophy. Yet there is
no substantial agreement as to precisely what is her definition of government, nor what her statements about government mean.
Questions and disagreements have arisen specifically over the meaning and ramifications of certain phrases in her
principal writings about government. In this article, I propose to identify those meaningings and ramifications, by way of
considering what Rand means by the term "a proper government." 
CONTROVERSIAL STATEMENTS AND PHRASES
In her essay "The Nature of
Government," Rand makes two statements which at least resemble definitions:
"A government is an institution that holds the exclusive power to enforce certain rules of social conduct in
a given geographic area." 
(G2) "A government is the means
of placing the retaliatory use of physical force under objective control -- i.e., under objectively defined laws."
Immediately preceding Statement (G2) in the text, is Rand's remark that protecting
people's rights under an objective code of rules "is the task of a government -- of a proper government
-- its basic task, its only moral justification and the reason why men do need a government." 
Here arise the disagreements over interpretations of her thesis. Is Statement (G2) actually the definition of "proper
government"? And what does "proper" mean in this context? Is it synonymous with "good"? "Good"
in what sense, and by what standard?
Or is Rand merely defining "government" in general?
Is she thus making "proper government" synonymous with "government" as she defines it, using
"proper" only as a superlative to distinguish things satisfying her definition from others which do not, yet which
are commonly called "government"? Or is the possibility of some other meaning intended? Does either statement,
(G1) or (G2) contain the necessary elements for a full and complete definition of the concept of "government"?
Consider the phrases "exclusive power" and "given geographical area" in Statement (G1). Does
exclusivity of power necessarily rule out the possibility of competing governments in the same geographical area? Does the
notion of competing governments in the same geographical area even make sense?
And what of the
phrase "is the means of placing" in Statement (G2)? Does this mean that governments actually have been used
this way? If so, then isn't that a wrong distinguishing characteristic to pick for the concept "government,"
since governments have always strongly tended toward violating rather than protecting men's rights?
Or were such historical examples justifiably called "governments"? If not, then Statement (G2)
seems to refer to nothing which has ever existed, and merely redefines the word "government" to make it inapplicable
to reality! It designates what we call a good government, none of which exist according to Objectivist-Libertarian
Or instead, does the phrase "is the means of placing" mean that government
can be used in this way, regardless of whether and to what extent it actually has been? If so, why is this
possible use regarded as the distinguishing characteristic, rather than, say, to provide public welfare?
The unresolved debate over these questions has caused a wide division among Libertarians and Objectivists. The issue
is: what constitutes a moral political system? All these questions arise because Rand's definition (if it is one) of "Government"
is much subtler than most people think.
As I have said, a central problem seems to be the precise
role and meaning of the concept "proper," as Rand employs it in this context. The solution to that problem, provided
below, will enable us to resolve the tangle of questions which have arisen, and will serve as the key to understanding her
view of government.
THE "PROPER" SOLUTION TO THE CONTROVERSIES
First, recall that Rand discusses the meaning of "proper" in a meta-ethical context in "the Objectivist
What standard determines what action is proper in this context?
The standard is the organism's life, or: that which is required for the organism's survival. 
A "proper action," then, is one which is consistent with the requirements of an
organism's survival, with its actual needs.
The concept "proper" is also
applicable to the two basic aspects of human action: the means -- the method or instrumentality by which the action
is carried out -- and the end the goal or purpose of the action. Thus, a "proper object of action" is one
of man's needs, that which will fulfill one of his survival requirements. And in relation to some given need of man, a
"proper means" is that which will help him to obtain that end.
Now consider the context
of Rand's discussion of government. She is asking: Why do people need a social-rule-enforcing institution? Why does a
person need an institution which sets and enforces limits upon his actions and those of others?
answer lies in man's basic social need: justice, the recognition and acceptance by others of his nature, of his
requisite conditions for proper survival, namely, freedom from the initiation of physical force by others, in other words,
the recognition of his rights. The means for fulfilling this need is, therefore, an institution which sets and enforces limits
upon everyone's actions, limits based upon the standard of man's rights. 
It is thus
clear that Rand is treating government as one sub-category within the wider class of social-rule-enforcing institutions.
(Observe that Statement (G1) does not assert that government is the only kind of social-rule-enforcing institution.
If she had intended to make such an assertion, the article "the" would have been appropriate instead of "an.")
The criterion by which Rand differentiates what she regards as government from the other members of the wider class
is whether or not people need them -- i.e., whether they to any extent serve as the means by which man can fulfill his basic
social need. If they do not help to fulfill that need then they are not classified by Rand as government.
Thus, contrary to the anarcho-capitalists' interpretation of Statement (G2) , a given institution may protect
very few rights and violate many, or pursue some other mixed policy, and still be consistent with what Statement (G2) says
An institution is thus classified as a government if it is the means of objectively
controlling retaliatory force (protecting rights) to at least some extent. Once we have ascertained that a given
institution does in fact meet this criterion for being a government, we then decide whether it is a good, bad, or mediocre
government, according to the overall extent to which it objectively controls retaliatory force (i.e., protects rights).
It is important to recognize that Rand, in developing her concept of government, is challenging common usage, denying
that certain things commonly called "government" are not essentially different from others in their ability
to fulfill man's basic needs. Thus, her expression "proper government" is actually a double-entendre, so to
In Rand's usage, the expression "proper government" is redundant,
bding synonymous with "government." The term "proper" in this sense serves only as a clarifier;
there is no such thing as an "improper government," in Rand's usage.
But in common
usage, the expression "proper government" is not redundant at all, but instead is elliptical, referring
to proper things commonly called "government". In this sense, the term "proper" more than clarifies:
it evaluates certain things commonly called "government" as being proper to (consistent with) man's
basic social need. In this sense there are "improper governments": things called "government´which
are not proper to man's basic social need.
It is now apparent that although neither
of Rand's aforemention statements about government is a definition, they each contain some of the necessary components
of a correct definition, and they jointly contain all the necessary components. Statement (G1) specifies the genus
of government, while Statement (G2) specifies the differentia of government.
basis, we can formulate a full and correct definition of government:
is the institution that holds the exclusive power to enforce certain rules of social conduct, at least some of which are
for the protection of men's rights (i.e., the objective control of retaliatory force), in a given geographical area.
We have now eliminated most of
the problems which obstruct an understanding of Rand's concept of government as being consistent and correct. The only
questions yet unresolved are those concerning the "geographical area" and "exclusive power" criteria in
Concerning the former criterion, it is possible that the given geographical area
(over which a single institution exclusively enforces certain rules) might be a number of spatially separated areas, rather
than one continuous region.  The tendency to visualize the "given geographical area" as one single, unbroken
expanse, can be corrected by asking: "Given -- by what?"
the "given-ness" is not to be totally arbitrary, restricted to continuous regions, then the "geographical area"
criterion must mean: the geographical area equivalent to the owned properties of those who choose
to become and/or remain governed by the same institution.
If interpreted in this way, the
criterion would still apply to the case where a government over a continuous country forbids secession and the citizens
choose to remain under its jurisdiction anyway, rather than revolt or move out. More important, it would allow for vastly
more possibilities, all properly referred to as government.
For instance, there might be a pattern
of spatially separate yet not too distant governed areas forming a patchwork. (Even in the United States, where we are forced
to support the Federal, State, and local governments, that is somewhat the case, and other private agencies perform governmental
functions in the same "patchwork" pattern.
But then how does this affect the latter
criterion? In such a free, moral society, wouldn't various governments compete with one another? In one misleading
sense, yes; they vie for the job of protecting their prospective citizens'/supporters'/customers' rights and perhaps
also enforcing other rules.
However, once the job is assigned, the geographical area
of jurisdiction is set, regardless of how temporary that arrangement may be. The competition is part of the process of seeking
citizens to govern, not part of the process of governing per se. That is, only one agency actually governs at a given
time and in a given respect. 
Thus, insofar as they are serving as governments, rather than
merely seeking to become governments, there is no such thing as competing governments in a given geographical area.
The notion is simply incoherent. 
In fairness, this is not what the anarcho-capitalists mean
to say, if I understand them correctly. They maintain, rather, that institutions in the same general location can bid to become
the government over two different geographical areas within this location, no matter how intertwined those areas may become
due to the location of their customers' properties. Thus, they are merely describing two or more governments coexisting
much as do the U.S. and Canada today, only in a somewhat more complicated geographical pattern.
is not to deny the possibility of "overlapping or shared jurisdiction" between governments, in which, for instance,
police officers from each government would cooperate and collaborate in apprehending criminals. Such cases, however, are examples
of a higher-level governmental structure under which (part or all of) two or more other governments are subsumed.
(This does not preclude two governments, in taking joint action, temporarily merging into one, which itself has exclusive
power of enforcement.)
There have been errors on both sides of the dispute. While many of the
anarchists have been trying to project what geographical form the coexistence of different governments might take, if enough
coercive barriers were removed, many of the limited governmentalists have been blind to that view, instead concentrating on
flaws in the form in which the anarchists have expressed their projection.
A famous example
of this error, unfortunately, is Rand's ill-considered example of Mr. Smith, a customer of Government A and his next-door
neighbor, Mr. Jones, a customer of Government B.  What, after all, is the line between Mr. Smith's property
and Mr. Jones' property in such a case -- if not example equivalent to the international border, say, between the U.S.
and Canada? (I invite the reader to draw his own conclusions as to whether it is impossible or even unlikely that Governments
A and B could work out mutually satisfactory methods for settling disputes between their citizens.)
is finally possible to lay the entire archy-anarchy debate to rest. As Rand has maintained, anarchy is "a naive
floating abstraction.´ The system the anarcho-capitalists propose, however, is not anarchy, nor is it competing
governments in the same geographical area -- it is merely a system of limited government(s), with the potential of being consistently
moral. If both sides would only realize this, they could discard all of the false labels and work together of one mind and
accord, to bring the ideal of such a governmental system into existence.
NOTES AND REFERENCES
 The views expressed in this paper are based strictly upon my own understanding and interpretation of Rand's
(and others') views, and are not to be construed as having her (or their) sanction or approval. The reader may find it
helpful to consider this issue in light of the fallacy of the Frozen Abstraction, first identified by Ayn Rand in her
essay "Collectivized Ethics," included in The Virtue of Selfishness. For details see the relevant
section of my essay, "On the Fine Art of Thawing Out Frozen Abstractions."
 Ayn Rand, The Virtue of Selfishness (Signet Books, New American
Library, 1964), p. 107.
 Op. cit., p. 109.
 Op. cit., p. 16.
 Op. cit., pp. 108-109.
 See L.
A. Rollins, "Some Brief Comments and Questions about Machan's Governmentalism," Invictus No. 11, p.
20; also, Ronn Neff, "A Note on Machan's Anarchism,"Individualist, March 1971, pp. 11-12.
 T. R. Machan, "A Note on Neff's Anarchism," Reason, January 1971, p. 19.
 Rand, op. cit., pp. 112-113.