Abstract
The duality theory of the Monge–Kantorovich transport problem is investigated in an abstract measure theoretic framework. Let \((\mathcal {X},\mathcal {F},\mu )\) and \((\mathcal {Y},\mathcal {G},\nu )\) be any probability spaces and \(c:\mathcal {X}\times \mathcal {Y}\rightarrow \mathbb {R}\) a measurable cost function such that \(f_1+g_1\le c\le f_2+g_2\) for some \(f_1,\,f_2\in L_1(\mu )\) and \(g_1,\,g_2\in L_1(\nu )\). Define \(\alpha (c)=\inf _P\int c\,dP\) and \(\alpha ^*(c)=\sup _P\int c\,dP\), where \(\inf \) and \(\sup \) are over the probabilities P on \(\mathcal {F}\otimes \mathcal {G}\) with marginals \(\mu \) and \(\nu \). Some duality theorems for \(\alpha (c)\) and \(\alpha ^*(c)\), not requiring \(\mu \) or \(\nu \) to be perfect, are proved. As an example, suppose \(\mathcal {X}\) and \(\mathcal {Y}\) are metric spaces and \(\mu \) is separable. Then, duality holds for \(\alpha (c)\) (for \(\alpha ^*(c)\)) provided c is upper-semicontinuous (lower-semicontinuous). Moreover, duality holds for both \(\alpha (c)\) and \(\alpha ^*(c)\) if the maps \(x\mapsto c(x,y)\) and \(y\mapsto c(x,y)\) are continuous, or if c is bounded and \(x\mapsto c(x,y)\) is continuous. This improves the existing results in Ramachandran and Ruschendorf (Probab Theory Relat Fields 101:311–319, 1995) if c satisfies the quoted conditions and the cardinalities of \(\mathcal {X}\) and \(\mathcal {Y}\) do not exceed the continuum.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Throughout, \((\mathcal {X},\mathcal {F},\mu )\) and \((\mathcal {Y},\mathcal {G},\nu )\) are probability spaces and
is the product \(\sigma \)-field on \(\mathcal {X}\times \mathcal {Y}\). Further, \(\Gamma (\mu ,\nu )\) is the collection of probability measures P on \(\mathcal {H}\) with marginals \(\mu \) and \(\nu \), namely
For any probability space \((\Omega ,\mathcal {A},Q)\), we write \(L_1(Q)\) to denote the class of \(\mathcal {A}\)-measurable and Q-integrable functions \(\phi :\Omega \rightarrow \mathbb {R}\) (without identifying maps which agree Q-a.s.). We also write \(Q(\phi )=\int \phi \,dQ\) for \(\phi \in L_1(Q)\).
With a slight abuse of notation, for any maps \(f:\mathcal {X}\rightarrow \mathbb {R}\) and \(g:\mathcal {Y}\rightarrow \mathbb {R}\), we still denote by f and g the functions on \(\mathcal {X}\times \mathcal {Y}\) given by \((x,y)\mapsto f(x)\) and \((x,y)\mapsto g(y)\). Thus, \(f+g\) is the map on \(\mathcal {X}\times \mathcal {Y}\) defined as
In this notation, we let
Let \(c:\mathcal {X}\times \mathcal {Y}\rightarrow \mathbb {R}\) be an \(\mathcal {H}\)-measurable function satisfying
For such a c, we define
It is not hard to see that
A duality theorem (for both \(\alpha (c)\) and \(\alpha ^*(c)\)) is the assertion that
Indeed, duality theorems arise in a plenty of frameworks. The main one is possibly mass transportation, where c(x, y) is regarded as the cost for moving a unit of good from \(x\in \mathcal {X}\) into \(y\in \mathcal {Y}\). However, duality results play a role even in risk theory, optimization problems and dependence modeling. See, e.g., [1, 3,4,5,6, 9, 11,12,13, 16, 17] and references therein.
Starting from Kantorovich himself [8], there is a long line of research on duality theorems; see again [3, 4, 17] and references therein. To our knowledge, under the present assumptions on c, the best result is due to Ramachandran and Ruschendorf [14]. According to the latter, one obtains both \(\alpha (c)=\beta (c)\) and \(\alpha ^*(c)=\beta ^*(c)\) provided c is \(\mathcal {H}\)-measurable, it satisfies condition (1), and at least one between \(\mu \) and \(\nu \) is perfect.
Now, some form of condition (1) cannot be dispensed, while removing measurability leads to involve inner and outer measures; see [9, Section 2]. Instead, whether the perfectness assumption can be dropped is still an open problem. Thus, if c is measurable and meets (1) but \(\mu \) and \(\nu \) are both non-perfect, it is currently unknown whether condition (2) is true or false. See points (2)–(3), page 355, of [15].
This paper provides duality theorems not requiring perfectness.
Suppose \(\mathcal {X}\) and \(\mathcal {Y}\) are metric spaces and \(\mathcal {F}\) and \(\mathcal {G}\) the Borel \(\sigma \)-fields. Then, condition (2) is shown to be true if at least one of \(\mu \) and \(\nu \) is separable, c meets (1) and all the c-sections are continuous. Or else, condition (2) holds if \(\mu \) and \(\nu \) are both separable, c is bounded and measurable, and at least one of the c-sections is continuous. These results improve [14] when c satisfies the quoted assumptions and the cardinalities of \(\mathcal {X}\) and \(\mathcal {Y}\) do not exceed the continuum. Under the latter condition, in fact, a perfect probability measure is separable but not conversely. Note also that if \(\mathcal {X}\) and \(\mathcal {Y}\) are separable metric spaces (so that separability of \(\mu \) and \(\nu \) is automatic) the scope of our results is to replace assumptions on \(\mu \) or \(\nu \) (required by [14]) with assumptions on c.
Various conditions for \(\alpha (c)=\beta (c)\) or \(\alpha ^*(c)=\beta ^*(c)\), but not necessarily for both, are given as well. For instance, if c meets (1) and at least one of \(\mu \) and \(\nu \) is separable, then \(\alpha ^*(c)=\beta ^*(c)\) or \(\alpha (c)=\beta (c)\) provided c is lower- or upper-semicontinuous. As another example, \(\alpha ^*(1_H)=\beta ^*(1_H)\) if \(H=\cup _n (A_n\times B_n)\) with \(A_n\in \mathcal {F}\) and \(B_n\in \mathcal {G}\). Further, \(\alpha (1_H)=\beta (1_H)\) if \(\mu (\limsup _nA_n)=0\) or \(\nu (\limsup _nB_n)=0\). Without some extra conditions, however, we do not know whether \(\alpha (1_H)=\beta (1_H)\).
2 Preliminaries
For any topological space S, the Borel \(\sigma \)-field on S is denoted by \(\mathcal {B}(S)\).
Let \((\Omega ,\mathcal {A},Q)\) be a probability space. Then, Q is perfect if, for any \(\mathcal {A}\)-measurable \(\phi :\Omega \rightarrow \mathbb {R}\), there is \(B\in \mathcal {B}(\mathbb {R})\) such that \(B\subset \phi (\Omega )\) and \(Q(\phi \in B)=1\).
An important special case is \(\Omega \) a metric space and \(\mathcal {A}=\mathcal {B}(\Omega )\). In that case, Q is separable if \(Q(A)=1\) for some separable \(A\in \mathcal {A}\) and Q is tight if \(Q(A)=1\) for some \(\sigma \)-compact \(A\in \mathcal {A}\). Clearly, tightness implies separability but not conversely. Furthermore, tightness is equivalent to perfectness provided \(\Omega \) satisfies the following condition:
-
The power set of \(\Omega \) does not support any 0–1-valued probability measure T such that \(T\{\omega \}=0\) for each \(\omega \in \Omega \);
see [10, Theorem 3.2].
Two remarks are in order. First, the above condition on \(\Omega \) is automatically true if card\((\Omega )\le \,\)card\((\mathbb {R})\). Thus, perfectness implies separability, but not conversely, if card\((\Omega )\le \,\)card\((\mathbb {R})\) (in particular, if \(\Omega \) is a separable metric space). Second, it is consistent with the usual axioms of set theory (ZFC) that, for any metric space \(\Omega \), any probability measure on \(\mathcal {B}(\Omega )\) is separable.
Note also that a simple example of non-perfect probability measure is any non-tight probability measure on the Borel sets of a separable metric space. For instance, take Q the outer Lebesgue measure on \(\mathcal {B}(\Omega )\), where \(\Omega \) is a subset of [0, 1] with outer Lebesgue measure 1 and inner Lebesgue measure 0. Then, Q is not perfect.
Let us come back to duality theorems. Define
and note that
Thus, to get condition (2), it suffices to show \(\alpha (c)=\beta (c)\) under some conditions which hold true for both c and \(-c\).
Two preliminary lemmas are needed. The first is inspired to [7, Lemma 1].
Lemma 1
Let \(c\in M\). Then, \(\beta ^*(c)=\lim _n\beta ^*(c_n)\) whenever \((c_n)\subset M\) is an increasing sequence such that \(c_n\uparrow c\) pointwise.
Proof
We first suppose \(0\le c_n\le c\le k\) for some integer k. Under this assumption, for each n, there is \(f_n+ g_n\in L\) such that
see, e.g., [9, Lemma 1.8].
Since the sequences \((f_n)\) and \((g_n)\) are uniformly bounded, there are \(f\in L_1(\mu )\), \(g\in L_1(\nu )\) and a subsequence \((m_n)\) such that
In turn, this implies the existence of a sequence \((\phi _n,\psi _n)\) such that \(\phi _n\rightarrow f\) in \(L_1(\mu )\), \(\psi _n\rightarrow g\) in \(L_1(\nu )\) and \((\phi _n,\psi _n)\) is a convex combination of \(\{(f_{m_j},g_{m_j}):j\ge n\}\) for each n. By taking a further subsequence, it can be also assumed \(\mu (\phi _n\rightarrow f)=\nu (\psi _n\rightarrow g)=1\). Since \((c_n)\) is increasing, \(\phi _n+\psi _n\ge c_{m_n}\). Hence, after modifying f and g on null sets, one obtains \(f+ g\ge c\). On noting that \((\beta ^*(c_n))\) is a monotone sequence, it follows that
This concludes the proof if \(0\le c_n\le c\le k\). To deal with the general case, fix \(p+q\in L\) such that \(p+q\le c_1\) and define \(b_n=c_n-(p+q)\) and \(b=c-(p+q)\). Then, \(0\le b_n\le b\). Further, since \(\beta ^*(h+p+q)=\beta ^*(h)+\mu (p)+\nu (q)\) for each \(h\in M\), it suffices to show that \(\beta ^*(b)=\lim _n\beta ^*(b_n)\).
Given k, take \(f_k+ g_k\in L\) such that
Take also \(f+ g\in L\) such that \(f+ g\ge b\) and note that
Similarly, \(g\,1_{\{f>k\}}\le g\,1_{\{g>k\}}+f\,1_{\{f>k\}}\). Hence,
Since \(f_k+ g_k+3f\,1_{\{f>k\}}+3g\,1_{\{g>k\}}\) belongs to L, it follows that
Fix \(\epsilon >0\) and take k such that \((1/k)+3\mu \bigl [f\,1_{\{f>k\}}\bigr ]+3\nu \bigl [g\,1_{\{g>k\}}\bigr ]<\epsilon \). By what already proved, \(\beta ^*(b\wedge 2k)=\lim _n\beta ^*(b_n\wedge 2k)\). Therefore,
This concludes the proof. \(\square \)
In the second lemma, and in the rest of the paper, we write \(\alpha (H)=\alpha (1_H)\) whenever \(H\in \mathcal {H}\). The same notation is adopted for \(\beta \), \(\alpha ^*\) and \(\beta ^*\).
Lemma 2
Let \(c\in M\). Then, condition (2) holds provided \(\alpha (H)=\beta (H)\) for each \(H\in \mathcal {H}\).
Proof
It suffices to show \(\alpha (c)=\beta (c)\). To this end, we first note that \(\beta (c)\) is attained, i.e., \(\beta (c)=\mu (f_1)+\nu (g_1)\) for some \(f_1+ g_1\in L\) such that \(f_1+ g_1\le c\); see [14, Proposition 3]. Define \(h=c-(f_1+g_1)\) and fix \(t>1\) and \(P\in \Gamma (\mu ,\nu )\). Then,
Take \(f_2+ g_2\in L\) such that \(f_2+ g_2\ge c\) and define
Since \(h\le f+ g\),
Arguing as in the proof of Lemma 1,
Hence,
Next, by Theorem 2.1.1 and Remark 2.1.2(b) of [13], there is a finitely additive probability Q on \(\mathcal {H}\), with marginals \(\mu \) and \(\nu \), such that \(Q(c)=\beta (c)\). Since Q has marginals \(\mu \) and \(\nu \), then \(\beta (H)\le Q(H)\) for all \(H\in \mathcal {H}\) and
Finally, since \(h\ge 0\) and \(\alpha (H)=\beta (H)\) for all \(H\in \mathcal {H}\), one obtains
Hence, there is \(P_t\in \Gamma (\mu ,\nu )\) such that \(P_t(h>t^{-1})<t^{-2}\). It follows that
Since \(1/t+\mu \bigl [f\,1_{\{f>t\}}\bigr ]+\nu \bigl [g\,1_{\{g>t\}}\bigr ]\rightarrow 0\) as \(t\rightarrow \infty \), this concludes the proof. \(\square \)
3 Duality Theorems Without Perfectness
It is convenient to distinguish two cases.
3.1 The Abstract Case
Theorem 3
Let \(c\in M\). Then, condition (2) holds provided
- (*):
-
For each \(\epsilon >0\), there is a countable partition \(\{A_0,A_1,\ldots \}\subset \mathcal {F}\) of \(\mathcal {X}\) such that \(\mu (A_0)=0\) and
$$\begin{aligned} \sup _{y\in \mathcal {Y}}\,|c(x,y)-c(z,y)|\le \epsilon \quad \text {whenever }x,\,z\in A_i\text { and }i>0. \end{aligned}$$
Proof
Again, it suffices to show \(\alpha (c)=\beta (c)\). Given \(\epsilon >0\), fix a point \(x_i\in A_i\) for each \(i>0\), and define
Let \(\Gamma (\mu _0,\nu )\) be the set of probability measures on \(\mathcal {F}_0\otimes \mathcal {G}\) with marginals \(\mu _0\) and \(\nu \).
Take \(f_1+ g_1\in L\) and \(f_2+ g_2\in L\) such that \(f_1+ g_1\le c\le f_2+ g_2\). Since \(|c-c_0|\le \epsilon \), then \(f_1+ g_1-\epsilon \le c_0\le f_2+ g_2+\epsilon \). Further, \(\sup _{A_i}f_1<+\infty \) and \(\inf _{A_i}f_2>-\infty \) for each \(i>0\). Define
Then, \(\phi _1,\,\phi _2\in L_1(\mu _0)\) and
Because of such inequality and since \(c_0\) is \(\mathcal {F}_0\otimes \mathcal {G}\)-measurable, one can define
where \(\sup \) is over the pairs (f, g) such that \(f\in L_1(\mu _0)\), \(g\in L_1(\nu )\) and \(f+ g\le c_0\).
Since \(\mu _0\) is an atomic probability measure, then \(\mu _0\) is perfect, which in turn implies \(\alpha _0=\beta _0\). Since \(|c-c_0|\le \epsilon \), then \(\beta _0\le \beta (c)+\epsilon \). Hence, there is \(T\in \Gamma (\mu _0,\nu )\) such that
If T can be extended to a probability measure \(P\in \Gamma (\mu ,\nu )\), then
Thus, to conclude the proof, it suffices to show that T can be actually extended to a probability measure \(P\in \Gamma (\mu ,\nu )\).
For each i with \(\mu (A_i)>0\), define
where \(A\in \mathcal {F}\) and \(B\in \mathcal {G}\). Define also
where \(\mu _i\times \nu _i\) is the product measure of \(\mu _i\) and \(\nu _i\) (so that \(\mu _i\times \nu _i\) is a probability measure on \(\mathcal {H}\)). It is straightforward to see that \(P\in \Gamma (\mu ,\nu )\). Fix \(A\in \mathcal {F}_0\) and \(B\in \mathcal {G}\). For \(i>0\), either \(A\cap A_i=\emptyset \) or \(A\cap A_i=A_i\), so that
Therefore, \(P=T\) on \(\mathcal {F}_0\otimes \mathcal {G}\). \(\square \)
In Theorem 3, clearly, the roles of \(\mu \) and \(\nu \) can be interchanged. Accordingly, condition (*) can be replaced by
- (**):
-
For each \(\epsilon >0\), there is a countable partition \(\{B_0,B_1,\ldots \}\subset \mathcal {G}\) of \(\mathcal {Y}\) such that \(\nu (B_0)=0\) and
$$\begin{aligned} \sup _{x\in \mathcal {X}}\,|c(x,y)-c(x,z)|\le \epsilon \quad \text {whenever }y,\,z\in B_i\text { and }i>0. \end{aligned}$$
As an example, condition (*) holds (with \(A_0=\emptyset \)) if \(\mathcal {X}\) is a separable metric space and the function \(x\mapsto c(x,y)\) is Lipschitz uniformly with respect to y, i.e.,
where \(u>0\) is a constant and d the distance on \(\mathcal {X}\). Fix in fact \(\epsilon >0\). Because of separability, \(\mathcal {X}\) can be partitioned into sets \(A_1,A_2,\ldots \) whose diameter is less than \(\epsilon /u\). Hence, condition (*) follows trivially from (3). Similarly, condition (**) holds if \(\mathcal {Y}\) is a separable metric space and \(y\mapsto c(x,y)\) is Lipschitz uniformly with respect to x. Further, as shown in the proof of Theorem 7, separability of \(\mathcal {X}\) (of \(\mathcal {Y}\)) can be weakened into separability of \(\mu \) (of \(\nu \)).
Another example is the following. Let \(\mathcal {R}\) be the field of subsets of \(\mathcal {X}\times \mathcal {Y}\) generated by the measurable rectangles. Each \(R\in \mathcal {R}\) can be written as \(R=\cup _{i=1}^n(A_i\times B_i)\) for some \(n\ge 1\) and \(A_i\in \mathcal {F}\), \(B_i\in \mathcal {G}\) such that \(A_i\cap A_j=\emptyset \) for \(i\ne j\). Thus, when \(c=1_R\) with \(R\in \mathcal {R}\), condition (*) is trivially true and Theorem 3 yields \(\alpha (R)=\beta (R)\) and \(\alpha ^*(R)=\beta ^*(R)\). We next prove duality of certain sets related to \(\mathcal {R}\).
Theorem 4
Let \(H=\cup _nR_n\) and \(K=\cap _nR_n\) where \(R_n\in \mathcal {R}\) for each n. Then,
In addition, \(\alpha (H)=\beta (H)\) provided H can be written as \(H=\cup _n (A_n\times B_n)\) with \(A_n\in \mathcal {F}\), \(B_n\in \mathcal {G}\), and
(Here, \(\limsup _nA_n=\cap _n\cup _{j>n}A_j\) and \(\limsup _nB_n=\cap _n\cup _{j>n}B_j\).)
Proof
Let \(H_n=\cup _{i=1}^nR_i\). Since \(\alpha ^*(H_n)=\beta ^*(H_n)\), Lemma 1 implies
Thus, \(\alpha ^*\) and \(\beta ^*\) agree on countable unions of elements of \(\mathcal {R}\). Since \(R_n^c\in \mathcal {R}\), this implies
Next, suppose \(H=\cup _n (A_n\times B_n)\) and \(\mu (\limsup _nA_n)=0\). Let \(V_n=\cup _{i=1}^n(A_i\times B_i)\). Given \(\epsilon >0\), take \(n\ge 1\) such that \(\mu \bigl (\cup _{i>n}A_i\bigr )<\epsilon \), and then, take \(P\in \Gamma (\mu ,\nu )\) satisfying \(P(V_n)<\alpha (V_n)+\epsilon \). Since \(\alpha (V_n)=\beta (V_n)\), one obtains
The proof is exactly the same if \(\nu (\limsup _nB_n)=0\). \(\square \)
Because of Theorem 4, a (classical) question raised by Arveson [2] admits a positive answer for countable unions of measurable rectangles.
Arveson’s problem If \(H\in \mathcal {H}\) satisfies \(P(H)=0\) for all \(P\in \Gamma (\mu ,\nu )\), are there \(A\in \mathcal {F}\) and \(B\in \mathcal {G}\) such that \(\mu (A)=\nu (B)=0\) and \(H\subset (A\times \mathcal {Y})\cup (\mathcal {X}\times B)\) ?
Indeed, it is not hard to see that \(\beta ^*(H)=\mu (A)+\nu (B)\) for some \(A\in \mathcal {F}\) and \(B\in \mathcal {G}\) with \(H\subset (A\times \mathcal {Y})\cup (\mathcal {X}\times B)\); see, e.g., [7, Lemma 1]. If H is a countable union of measurable rectangles, Theorem 4 implies \(\beta ^*(H)=\alpha ^*(H)=0\) so that \(\mu (A)=\nu (B)=0\).
In addition, exploiting Theorem 4, duality for \(H=\cup _n (A_n\times B_n)\) can be obtained under various conditions. One such condition is \(\mu (\limsup _nA_n)=0\) or \(\nu (\limsup _nB_n)=0\). A similar condition is that \(H^c\) is also a countable union of measurable rectangles. In this case, in fact, \(\alpha ^*(H^c)=\beta ^*(H^c)\) or equivalently \(\alpha (H)=\beta (H)\). A last condition is
Define in fact \(P_n(U)=\mu \bigl \{x:(x,\phi _n(x))\in U\bigr \}\) for each \(n\ge 1\) and \(U\in \mathcal {H}\). Then, \(P_n\in \Gamma (\mu ,\nu )\) and \(\alpha (H)\le P_n(H)< 1/n\). Thus, \(\alpha (H)=0\), which in turn implies \(\alpha (H)=\beta (H)\). Here is a simple example.
Example 5
Suppose \((\mathcal {X},\mathcal {F})=(\mathcal {Y},\mathcal {G})\) and \(\mu =\nu \), with \(\mathcal {X}\) a separable metric space and \(\mathcal {F}=\mathcal {B}(\mathcal {X})\). (Up to some technicalities, separability of \(\mathcal {X}\) could be weakened into separability of \(\mu \).) Let \(\Delta =\{(x,x):x\in \mathcal {X}\}\) be the diagonal and H a countable union of measurable rectangles. Then, duality holds for \(H\cap \Delta ^c\), and it holds for \(H\cap \Delta \) provided \(\mu \) vanishes on singletons. In fact, \(H\cap \Delta ^c\) is a countable union of measurable rectangles and \(\mu \bigl \{x:(x,x)\in H\cap \Delta ^c\bigr \}=0\). Letting \(\phi _n(x)=x\), Theorem 4 and condition (4) yield
To deal with \(H\cap \Delta \), suppose \(\mu \) null on singletons and define \(P_1=\mu \times \mu \) and \(P_2(U)=\mu \bigl \{x:(x,x)\in U\bigr \}\) for each \(U\in \mathcal {H}\). Then, \(P_1,\,P_2\in \Gamma (\mu ,\mu )\). Since \(\mu \) is null on singletons, \(\alpha (H\cap \Delta )\le P_1(H\cap \Delta )\le P_1(\Delta )=0\), which in turn implies \(\alpha (H\cap \Delta )=\beta (H\cap \Delta )\). Finally, writing H as \(H=\cup _n(A_n\times B_n)\), one obtains
We close this subsection with two remarks. The first (stated as a lemma) suggests a possible strategy for proving a general duality theorem.
Lemma 6
Let \(\mathcal {H}_0=\bigl \{H\in \mathcal {H}:\alpha (H)=\beta (H)\) and \(\alpha ^*(H)=\beta ^*(H)\bigr \}\). Then, condition (2) holds for each \(c\in M\) if and only if
Proof
By Lemma 2, it suffices to show that \(\mathcal {H}_0=\mathcal {H}\). In turn, since \(\mathcal {R}\subset \mathcal {H}_0\), it suffices to see that \(\mathcal {H}_0\) is a monotone class. Also, since \(\mathcal {H}_0\) is closed under complements, it is enough to prove that \(H\in \mathcal {H}_0\) provided H is the union of an increasing sequence of elements of \(\mathcal {H}_0\). Let \(H=\cup _n H_n\) where \(H_n\in \mathcal {H}_0\) and \(H_n\subset H_{n+1}\) for each n. For such H, arguing as in the proof of Theorem 4, one obtains \(\alpha ^*(H)=\beta ^*(H)\). Thus, under (5), \(\mathcal {H}_0\) is actually a monotone class. \(\square \)
The second remark briefly compares the arguments underlying Theorem 3 and the usual duality theorems. The latter are summarized into the result by Ramachandran and Ruschendorf [14].
For definiteness, we aim to prove \(\alpha (c)=\beta (c)\). By (1) and since \(\beta (c)\) is attained, it can be assumed \(c\ge 0\) and \(\beta (c)=0\). As noted in the proof of Lemma 2, there is a finitely additive probability Q on \(\mathcal {H}\), with marginals \(\mu \) and \(\nu \), satisfying \(Q(c)=\beta (c)\). Since \(c\ge 0\) and \(\beta (c)=0\), it must be \(Q(c>\epsilon )=0\) for each \(\epsilon >0\). A basic intuition in [14] is that if one of \(\mu \) and \(\nu \) is perfect, then Q is \(\sigma \)-additive on \(\mathcal {R}\); see [13, Theorem 2.1.3] and recall that \(\mathcal {R}\) is the field generated by the measurable rectangles. Hence, there is \(P\in \Gamma (\mu ,\nu )\) such that \(P=Q\) on \(\mathcal {R}\). With such a P, one obtains
provided \(R_i\in \mathcal {R}\) for all i. Hence, \(P(c>\epsilon )\le Q(c>\epsilon )=0\) if the set \(\{c>\epsilon \}\) is a countable union of measurable rectangles. Up to some technicalities, suitable versions of this argument work even if \(\{c>\epsilon \}\) fails to be a countable union of measurable rectangles. This provides a rough sketch of the proof of \(\alpha (c)=\beta (c)\) under the assumption that one of \(\mu \) and \(\nu \) is perfect. We now turn to Theorem 3. Here, instead of proving that Q is \(\sigma \)-additive on \(\mathcal {R}\), one requires that c can be suitably approximated by \(\mathcal {R}\)-simple functions. For instance, conditions (*)–(**) are trivially true if c is the uniform limit of a sequence of \(\mathcal {R}\)-simple functions, and in this case, no assumptions on \(\mu \) or \(\nu \) are needed. Apparently, conditions (*)–(**) are too restrictive to be useful in real problems. Instead, they allow to get duality in various situations, including Theorem 4, Example 5 and the results in the next subsection.
3.2 The Metric Case
In this subsection, \(\mathcal {X}\) and \(\mathcal {Y}\) are metric spaces, \(\mathcal {F}=\mathcal {B}(\mathcal {X})\) and \(\mathcal {G}=\mathcal {B}(\mathcal {Y})\). The sections of c are the functions \(x\mapsto c(x,y)\) and \(y\mapsto c(x,y)\), with y fixed in the first map and x fixed in the second.
A remark is in order. All claims made so far are still valid, even if c is not \(\mathcal {H}\)-measurable, provided \(c\,1_{A\times B}\) is \(\mathcal {H}\)-measurable for some \(A\in \mathcal {F}\) and \(B\in \mathcal {G}\) with \(\mu (A)=\nu (B)=1\). In fact, \(\alpha (c)=\alpha (c\,1_{A\times B})\) whenever \(\alpha (c)\) is defined in the obvious way, i.e.,
Similarly, \(\alpha ^*(c)=\alpha ^*(c\,1_{A\times B})\), \(\beta (c)=\beta (c\,1_{A\times B})\) and \(\beta ^*(c)=\beta ^*(c\,1_{A\times B})\).
In the next result, \(c\,1_{A\times B}\) is actually \(\mathcal {H}\)-measurable for some \(A\in \mathcal {F}\) and \(B\in \mathcal {G}\) such that \(\mu (A)=\nu (B)=1\) (with possibly \(A=\mathcal {X}\) or \(B=\mathcal {Y}\)).
Theorem 7
Suppose c satisfies condition (1), the map \(x\mapsto c(x,y)\) is continuous for each \(y\in \mathcal {Y}\) and the map \(y\mapsto c(x,y)\) is \(\mathcal {G}\)-measurable for each \(x\in \mathcal {X}\). Then,
-
(i)
\(\alpha ^*(c)=\beta ^*(c)\) if c is bounded below and \(\mu \) is separable;
-
(ii)
\(\alpha (c)=\beta (c)\) if c is bounded above and \(\mu \) is separable;
-
(iii)
\(\alpha (c)=\beta (c)\) and \(\alpha ^*(c)=\beta ^*(c)\) if c is bounded and \(\mu \) is separable;
-
(iv)
\(\alpha (c)=\beta (c)\) and \(\alpha ^*(c)=\beta ^*(c)\) if all the sections of c are continuous and at least one of \(\mu \) and \(\nu \) is separable.
Proof
Since (ii) and (iii) are consequences of (i), it suffices to prove (i) and (iv).
Let \(\mu \) and c be as in (i). Since \(\mu \) is separable, there is a separable set \(A\in \mathcal {F}\) with \(\mu (A)=1\). Since \(x\mapsto c(x,y)\) is continuous, \(y\mapsto c(x,y)\) is Borel measurable and A is separable, the restriction of c on \(A\times \mathcal {Y}\) is measurable with respect to \(\mathcal {B}(A)\otimes \mathcal {B}(\mathcal {Y})\). Therefore, \(c\,1_A\) is \(\mathcal {H}\)-measurable.
Take a countable set \(D\subset A\) such that \(\overline{D}=\overline{A}\) and define
where \(n\ge 1\), \((x,y)\in \mathcal {X}\times \mathcal {Y}\) and d is the distance on \(\mathcal {X}\).
Since c is bounded below, \(c_n\) is real-valued, and a direct calculation shows that
Since D is countable, \(y\mapsto c_n(x,y)\) is Borel measurable. Hence, \(c_n\,1_A\) is \(\mathcal {H}\)-measurable. In addition, \(c_n\le c_{n+1}\) and \(c_n\) meets condition (1) (since c meets (1) and is bounded below). Finally, since \(x\mapsto c(x,y)\) is continuous, one obtains
Next, \(\overline{D}=\overline{A}\) implies \(A\subset \bigcup _{x\in D}B(x,\delta )\) for each \(\delta >0\), where \(B(x,\delta )\) is the \(\mathcal {X}\)-ball of radius \(\delta \) around x. Given \(n\ge 1\) and \(\epsilon >0\), it follows that A can be partitioned into sets \(A_1,A_2,\ldots \in \mathcal {F}\) whose diameter is less than \(\epsilon /n\). Hence, \(c_n\) meets condition (*) (with \(A_0=A^c\)) because of (6). By Lemma 1 and Theorem 3,
This concludes the proof of (i).
Let us turn to (iv). Suppose that all the c-sections are continuous. Since the sections of \(-c\) are continuous as well, it suffices to prove \(\alpha ^*(c)=\beta ^*(c)\). We first assume \(\mu \) separable.
By (1), there are \(\psi \in L_1(\mu )\) and \(g\in L_1(\nu )\) such that \(\psi +g\le c\). Define
and note that \(f\in L_1(\mu )\), \(f+g\le c\) and f is upper-semicontinuous. Define also
Again, condition (6) holds, \(c_n\) meets condition (1) (since \(g\le c_n\le c-f\)) and \(y\mapsto c_n(x,y)\) is Borel measurable. (It is in fact upper-semicontinuous.) On noting that \(x\mapsto c(x,y)-f(x)\) is lower-semicontinuous, it is not hard to see that \(c_n\uparrow c-f\) pointwise as \(n\rightarrow \infty \). Because of (6) and \(\mu \) separable, \(c_n\) meets condition (*). Thus,
Hence, \(\alpha ^*(c)=\beta ^*(c)\) if \(\mu \) is separable.
Finally, if \(\nu \) is separable, it suffices to let
where now g is upper-semicontinuous and \(\rho \) is the distance on \(\mathcal {Y}\). Arguing as above and using separability of \(\nu \), it follows that \(c_n\) meets condition (**) and \(c_n\uparrow c-g\) pointwise as \(n\rightarrow \infty \). Hence, \(\alpha ^*(c)=\beta ^*(c)\) and this concludes the proof. \(\square \)
Once again, the roles of \(\mu \) and \(\nu \) can be interchanged in Theorem 7.
Theorem 8
Suppose c satisfies condition (1), the map \(x\mapsto c(x,y)\) is \(\mathcal {F}\)-measurable for each \(y\in \mathcal {Y}\) and the map \(y\mapsto c(x,y)\) is continuous for each \(x\in \mathcal {X}\). Then,
-
(i)
\(\alpha ^*(c)=\beta ^*(c)\) if c is bounded below and \(\nu \) is separable;
-
(ii)
\(\alpha (c)=\beta (c)\) if c is bounded above and \(\nu \) is separable;
-
(iii)
\(\alpha (c)=\beta (c)\) and \(\alpha ^*(c)=\beta ^*(c)\) if c is bounded and \(\nu \) is separable.
Note that if \(\mu \) and \(\nu \) are both separable, then \(\alpha (c)=\beta (c)\) and \(\alpha ^*(c)=\beta ^*(c)\) provided c is bounded, \(\mathcal {H}\)-measurable, and at least one of the c-sections is continuous. Further, the argument underlying Theorems 7–8 yields other similar results. As an example, we state (without a proof) the following.
Theorem 9
Suppose c satisfies condition (1) and at least one of \(\mu \) and \(\nu \) is separable. Then, \(\alpha ^*(c)=\beta ^*(c)\) if c is lower-semicontinuous (with respect to the product topology on \(\mathcal {X}\times \mathcal {Y}\)) and \(\alpha (c)=\beta (c)\) if c is upper-semicontinuous.
Finally, we list some consequences of Theorems 7–9. Indeed, they unify and slightly improve some known results.
-
Theorems 7–8 improve [14], the result by Ramachandran and Ruschendorf, provided c satisfies some conditions and
$$\begin{aligned} \text {card}(\mathcal {X})\le \text {card}(\mathbb {R})\quad \text {and}\quad \text {card}(\mathcal {Y})\le \text {card}(\mathbb {R}). \end{aligned}$$Under such cardinality assumption, in fact, perfectness implies separability but not conversely; see Sect. 2.
-
As an example, suppose \(c\in M\) and \(\mathcal {X}\) and \(\mathcal {Y}\) are separable metric spaces (so that \(\mu \) and \(\nu \) are both separable and the cardinality assumption is satisfied). Then, Ramachandran and Ruschendorf [14] implies \(\alpha (c)=\beta (c)\) and \(\alpha ^*(c)=\beta ^*(c)\) provided at least one between \(\mu \) and \(\nu \) is perfect. Instead, Theorems 7–8 lead to the same conclusions whenever all the c-sections are continuous, or whenever c is bounded and at least one of the c-sections is continuous.
-
By Theorems 7–8, it is consistent with the usual axioms of set theory (ZFC) that condition (2) holds for every \(c\in M\) with continuous sections, or for every bounded \(c\in M\) with at least one continuous section. In fact, as noted in Sect. 2, it is consistent with ZFC that any Borel probability on any metric space is separable.
-
Let \(\mathcal {X}=\mathcal {Y}\) and \(c=d\), where d is the distance on \(\mathcal {X}\). Suppose d measurable with respect to \(\mathcal {B}(\mathcal {X})\otimes \mathcal {B}(\mathcal {X})\) and
$$\begin{aligned} \int d(x,x_0)\,\mu (dx)+\int d(x,x_0)\,\nu (dx)<\infty \quad \text {for some }x_0\in \mathcal {X}. \end{aligned}$$Then, \(\alpha (d)\) reduces to Wasserstein distance between \(\mu \) and \(\nu \), while \(\beta (d)\) can be written as
$$\begin{aligned} \beta (d)=\sup _f\,|\mu (f)-\nu (f)|, \end{aligned}$$where \(\sup \) is over the 1-Lipschitz functions \(f:\mathcal {X}\rightarrow \mathbb {R}\). In this case, it is well known that \(\alpha (d)=\beta (d)\) if \(\mathcal {X}\) is separable; see, e.g., [9, page 400]. This known fact is generalized by Theorems 7–9 under two respects: Separability of \(\mathcal {X}\) can be weakened into separability of at least one of \(\mu \) and \(\nu \), and d can be replaced by any upper-semicontinuous function or by any function with continuous sections.
-
By Theorem 9, Arveson’s question has a positive answer if H is open and one of \(\mu \) and \(\nu \) is separable.
References
Ambrosio, L., Gigli, N., Savare’, G.: Gradient Flows, 2nd edn. Birkhauser, Basel (2008)
Arveson, W.: Operator algebras and invariant subspaces. Ann. Math. 100, 433–533 (1974)
Beiglbock, M., Schachermayer, W.: Duality for Borel measurable cost functions. Trans. Am. Math. Soc. 363, 4203–4224 (2011)
Beiglbock, M., Leonard, C., Schachermayer, W.: A general duality theorem for the Monge–Kantorovich transport problem. Studia Math. 209, 151–167 (2012)
Berti, P., Pratelli, L., Rigo, P.: Two versions of the fundamental theorem of asset pricing. Electron. J. Probab. 20, 1–21 (2015)
Berti, P., Pratelli, L., Rigo, P., Spizzichino, F.: Equivalent or absolutely continuous probability measures with given marginals. Depend. Model. 3, 47–58 (2015)
Haydon, R., Shulman, V.: On a measure-theoretic problem of Arveson. Proc. Am. Math. Soc. 124, 497–503 (1996)
Kantorovich, L.: On the translocation of masses. C. R. (Doklady) Acad. Sci. URSS (N.S.) 37, 199–201 (1942)
Kellerer, H.G.: Duality theorems for marginal problems. Z. Wahrscheinlichkeitstheorie Verw. Geb. 67, 399–432 (1984)
Koumoullis, G.: On perfect measures. Trans. Am. Math. Soc. 264, 521–537 (1981)
Puccetti, G., Wang, R.: Extremal dependence concepts. Stat. Sci. 30, 485–517 (2015)
Puccetti, G., Rigo, P., Wang, B., Wang, R.: Centers of probability measures without the mean. J. Theor. Probab. 32, 1482–1501 (2019)
Rachev, S.T., Ruschendorf, L.: Mass Transportation Problems, Volume I: Theory. Springer, New York (1998)
Ramachandran, D., Ruschendorf, L.: A general duality theorem for marginal problems. Probab. Theory Relat. Fields 101, 311–319 (1995)
Ramachandran, D., Ruschendorf, L.: On the Monge–Kantorovitch duality theorem. Theory Probab. Appl. 45, 350–356 (2000)
Ruschendorf, L.: Mathematical Risk Analysis: Dependence, Risk Bounds, Optimal Allocations and Portfolios. Springer, Heidelberg (2013)
Villani, C.: Optimal Transport, Old and New. Springer, New York (2009)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Rigo, P. A Note on Duality Theorems in Mass Transportation. J Theor Probab 33, 2337–2350 (2020). https://doi.org/10.1007/s10959-019-00932-x
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10959-019-00932-x
Keywords
- Duality theorem
- Mass transportation
- Perfect probability measure
- Probability measure with given marginals
- Separable probability measure