- A Proofs Before turning to the proofs, we need to introduce notation. Following Agueh & Carlier (2011), we define the space of continuous functions with at most quadratic growth by Cq := (1 + k k2 2)Cb(Rd ) = f ∈ C(Rd ) : f 1 + k k2 is bounded , which will be equipped with the norm kfkq := sup y∈Rd |f(y)| 1 + kyk2 , where k k2 denotes the Euclidean norm in Rd and Cb(Rd ) denotes the space of all bounded and continuous functions. Throughout, we work in the closed subspace Cq,0 of Cq given by Cq,0 := (1 + k k2 2)C0(Rd ) = f ∈ C(Rd ) : lim kyk2→∞ f(y) 1 + kyk2 = 0 ,
Paper not yet in RePEc: Add citation now
- Abadie, A. (2019), Using synthetic controls: Feasibility, data requirements, and methodological aspects, Technical report, MIT.
Paper not yet in RePEc: Add citation now
Abadie, A. & Gardeazabal, J. (2003), ‘The economic costs of conflict: A case study of the basque country’, American economic review 93(1), 113–132.
- Abadie, A. & L’Hour, J. (2017), A penalized synthetic control estimator for disaggregated data, Technical report, MIT.
Paper not yet in RePEc: Add citation now
Abadie, A., Diamond, A. & Hainmueller, J. (2010), ‘Synthetic control methods for comparative case studies: Estimating the effect of california’s tobacco control program ’, Journal of the American statistical Association 105(490), 493–505.
- Abadie, A., Diamond, A. & Hainmueller, J. (2015), ‘Comparative politics and the synthetic control method’, American Journal of Political Science 59(2), 495–510.
Paper not yet in RePEc: Add citation now
- Agueh, M. & Carlier, G. (2011), ‘Barycenters in the wasserstein space’, SIAM Journal on Mathematical Analysis 43(2), 904 – 924.
Paper not yet in RePEc: Add citation now
- Alexandrov, A. (1951), ‘A theorem on triangles in a metric space and some of its applications’, Trudy Mat. Inst. Steklov. 38, 5–23.
Paper not yet in RePEc: Add citation now
Aliprantis, C. D. & Border, K. (2006), Infinite Dimensional Analysis: a hitchhiker’s guide, Springer Science & Business Media.
Allegretto, S., Dube, A., Reich, M. & Zipperer, B. (2017), ‘Credible research designs for minimum wage studies: A response to neumark, salas, and wascher’, ILR Review 70(3), 559–592.
- Altschuler, J., Weed, J. & Rigollet, P. (2017), Near-linear time approximation algorithms for optimal transport via sinkhorn iteration, in ‘Advances in Neural Information Processing Systems’, pp. 1964–1974.
Paper not yet in RePEc: Add citation now
- Ambrosio, L., Gigli, N. & SavareÃŒÂ, G. (2008), Gradient flows in metric spaces and in the space of probability measures, Springer Since & Business Media.
Paper not yet in RePEc: Add citation now
- Amjad, M., Shah, D. & Shen, D. (2018), ‘Robust synthetic control’, The Journal of Machine Learning Research 19(1), 802–852.
Paper not yet in RePEc: Add citation now
Arkhangelsky, D., Athey, S., Hirshberg, D. A., Imbens, G. W. & Wager, S. (2019), Synthetic difference in differences, Technical report, National Bureau of Economic Research.
- Athey, S. & Imbens, G. W. (2006), ‘Identification and inference in nonlinear difference-in-differences models’, Econometrica 74(2), 431–497.
Paper not yet in RePEc: Add citation now
Athey, S. & Imbens, G. W. (2018), Design-based analysis in difference-in-differences settings with staggered adoption, Technical report, National Bureau of Economic Research.
Athey, S., Bayati, M., Doudchenko, N., Imbens, G. & Khosravi, K. (2018), Matrix completion methods for causal panel data models, Technical report, National Bureau of Economic Research.
- Ben-Michael, E., Feller, A. & Rothstein, J. (2018), New perspectives on the synthetic control method, Technical report, UC Berkeley.
Paper not yet in RePEc: Add citation now
- Benamou, J.-D. & Brenier, Y. (2000), ‘A computational fluid mechanics solution to the monge-kantorovich mass transfer problem’, Numerische Mathematik 84(3), 375–393.
Paper not yet in RePEc: Add citation now
- Benamou, J.-D., Froese, B. D. & Oberman, A. M. (2014), ‘Numerical solution of the optimal transportation problem using the monge–ampère equation’, Journal of Computational Physics 260, 107–126.
Paper not yet in RePEc: Add citation now
- Borjas, G. J. (2017), ‘The wage impact of the marielitos: A reappraisal’, ILR Review 70(5), 1077–1110.
Paper not yet in RePEc: Add citation now
- CaÃŒÂrcamo, J., RodrıÌÂguez, L.-A. & Cuevas, A. (2019), ‘Directional differentiability for supremum-type functionals: statistical applications’, arXiv: 1902.01136 .
Paper not yet in RePEc: Add citation now
- Chartrand, R., Wohlberg, B., Vixie, K. & Bollt, E. (2009), ‘A gradient descent solution to the monge-kantorovich problem’, Applied Mathematical Sciences 3(22), 1071–1080.
Paper not yet in RePEc: Add citation now
- Chernozhukov, V., Galichon, A., Hallin, M. & Henry, M. (2017), ‘Monge–kantorovich depth, quantiles, ranks and signs’, The Annals of Statistics 45(1), 223–256.
Paper not yet in RePEc: Add citation now
- Chernozhukov, V., Wuthrich, K. & Zhu, Y. (2018), ‘Inference on average treatment effects in aggregate panel data settings’, arXiv:1812.10820 .
Paper not yet in RePEc: Add citation now
Colson, B., Marcotte, P. & Savard, G. (2007), ‘An overview of bilevel optimization’, Annals of operations research 153(1), 235–256.
Cunningham, S. & Shah, M. (2017), ‘Decriminalizing indoor prostitution: Implications for sexual violence and public health’, The Review of Economic Studies 85(3), 1683–1715.
- Cuturi, M. (2013), Sinkhorn distances: Lightspeed computation of optimal transport, in ‘Advances in neural information processing systems’, pp. 2292–2300.
Paper not yet in RePEc: Add citation now
- Cuturi, M. & Doucet, A. (2014), Fast computation of wasserstein barycenters, in ‘International Conference on Machine Learning’, pp. 685–693.
Paper not yet in RePEc: Add citation now
- del Barrio, E. & Loubes, J.-M. (2019), ‘Central limit theorems for empirical transportation cost in general dimension’, The Annals of Probability 47(2), 926–951.
Paper not yet in RePEc: Add citation now
Doudchenko, N. & Imbens, G. W. (2016), Balancing, regression, difference-indifferences and synthetic control methods: A synthesis, Technical report, National Bureau of Economic Research.
- Dümbgen, L. (1993), ‘On nondifferentiable functions and the bootstrap’, Probability Theory and Related Fields 95(1), 125–140.
Paper not yet in RePEc: Add citation now
- Dube, A. (2019), ‘Minimum wages and the distribution of family incomes’, American Economic Journal: Applied Economics 11(4), 268–304.
Paper not yet in RePEc: Add citation now
- Ekeland, I., Heckman, J. J. & Nesheim, L. (2004), ‘Identification and estimation of hedonic models’, Journal of political economy 112(S1), S60–S109.
Paper not yet in RePEc: Add citation now
- Fang, Z. & Santos, A. (2018), ‘Inference on directionally differentiable functions’, The Review of Economic Studies 86(1), 377–412.
Paper not yet in RePEc: Add citation now
- Feydy, J., SeÃŒÂjourneÃŒÂ, T., Vialard, F.-X., Amari, S.-i., Trouve, A. & PeyreÃŒÂ, G. (2019), Interpolating between optimal transport and mmd using sinkhorn divergences, in ‘The 22nd International Conference on Artificial Intelligence and Statistics’, pp. 2681–2690.
Paper not yet in RePEc: Add citation now
- Flamary, R. & Courty, N. (2017), ‘POT: python optimal transport library’.
Paper not yet in RePEc: Add citation now
- Genevay, A., Chizat, L., Bach, F., Cuturi, M. & PeyreÃŒÂ, G. (2019), Sample complexity of sinkhorn divergences, in K. Chaudhuri & M. Sugiyama, eds, ‘Proceedings of Machine Learning Research’, Vol. 89 of Proceedings of Machine Learning Research, PMLR, pp. 1574–1583.
Paper not yet in RePEc: Add citation now
- Genevay, A., Cuturi, M., PeyreÃŒÂ, G. & Bach, F. (2016), Stochastic optimization for large-scale optimal transport, in ‘Advances in neural information processing systems ’, pp. 3440–3448.
Paper not yet in RePEc: Add citation now
Gunsilius, F. F. & Schennach, S. M. (2019), Independent nonlinear component analysis. Cemmap working paper.
- Hayfield, T. & Racine, J. S. (2008), ‘Nonparametric econometrics: The np package’, Journal of Statistical Software 27(5).
Paper not yet in RePEc: Add citation now
- Hong, H. & Li, J. (2018), ‘The numerical delta method’, Journal of Econometrics 206(2), 379–394.
Paper not yet in RePEc: Add citation now
Imai, K. & Kim, I. S. (2019), ‘When should we use unit fixed effects regression models for causal inference with longitudinal data?’, American Journal of Political Science 63(2), 467–490.
Jardim, E., Long, M. C., Plotnick, R., Van Inwegen, E., Vigdor, J. & Wething, H. (2017), Minimum wage increases, wages, and low-wage employment: Evidence from seattle, Technical report, National Bureau of Economic Research.
- Johnson, W. B. & Lindenstrauss, J. (1984), ‘Extensions of lipschitz mappings into a hilbert space’, Contemporary mathematics 26(189-206), 1.
Paper not yet in RePEc: Add citation now
- Kendall, D. G., Barden, D., Carne, T. K. & Le, H. (2009), Shape and shape theory, Vol. 500, John Wiley & Sons.
Paper not yet in RePEc: Add citation now
- Kloeckner, B. (2010), ‘A geometric study of wasserstein spaces: Euclidean spaces’, Annali della Scuola Normale Superiore di Pisa-Classe di Scienze 9(2), 297–323.
Paper not yet in RePEc: Add citation now
- Lock, E. F. (2018), ‘Tensor-on-tensor regression’, Journal of Computational and Graphical Statistics 27(3), 638–647.
Paper not yet in RePEc: Add citation now
- Matzkin, R. L. (2003), ‘Nonparametric estimation of nonadditive random functions’, Econometrica 71(5), 1339–1375.
Paper not yet in RePEc: Add citation now
- Mazur, S. & Ulam, S. (1932), ‘Sur les transformations isomeÃŒÂtriques d’espaces vectoriels normeÃŒÂs’, CR Acad. Sci. Paris 194(946-948), 116.
Paper not yet in RePEc: Add citation now
- Mena, G. & Weed, J. (2019), Statistical bounds for entropic optimal transport: sample complexity and the central limit theorem. arXiv: 1905.11882.
Paper not yet in RePEc: Add citation now
- Neumark, D. & Wascher, W. (2017), ‘Reply to ?credible research designs for minimum wage studies?’, ILR Review 70(3), 593–609.
Paper not yet in RePEc: Add citation now
Newey, W. K. (1991), ‘Uniform convergence in probability and stochastic equicontinuity ’, Econometrica: Journal of the Econometric Society pp. 1161–1167.
- Newey, W. K. & McFadden, D. (1994), ‘Large sample estimation and hypothesis testing’, Handbook of econometrics 4, 2111–2245.
Paper not yet in RePEc: Add citation now
Peri, G. & Yasenov, V. (2019), ‘The labor market effects of a refugee wave synthetic control method meets the mariel boatlift’, Journal of Human Resources 54(2), 267– 309.
- PeyreÃŒÂ, G. & Cuturi, M. (2019), ‘Computational optimal transport’, Foundations and Trends in Machine Learning 11(5-6), 355–607.
Paper not yet in RePEc: Add citation now
- Proof of Proposition 1. Lemmas 2 and 3 have shown the existence of solutions to (3) under Assumption 1. Furthermore, Propositions 3.5 in Agueh & Carlier (2011) shows that a solution to (4) exists and is unique under Assumptions 1 and 2. Finally, Remark 3.9 in Agueh & Carlier (2011) shows that if condition (5) holds for some measure P, i.e. where ∇Õj is the optimal transport map between PYjt and P, then P coincides with the barycenter P(λ) for this given λ. Therefore, if there exists a λ∗ ∈ ∆J−1 such that condition (5) holds where ∇Õj is the optimal transport map between PYjt and PY1t , then PY1t coincides with the unique barycenter P(λ∗ ) for given λ∗
Paper not yet in RePEc: Add citation now
- Proof. The following standard mathematical argument implies that P2(Y) is compact in the weak topology (i.e. the topology with respect to the dual space of Cq(Y)). First, the Banach-Alaoglu theorem (Aliprantis & Border 2006, Theorem 6.21) implies that the closed unit ball in M2(Rd ) is compact in the weak∗ -topology, i.e. the topology with respect to the dual space of Cq,0(Rd ). Second, the cone of probability measures in M2(Rd ) is closed, so that the intersection P2(Rd ) is compact in the weak∗ -topology. Now since we consider a compact subset Yt ⊂ Rd , it follows that Cq(Y) = C0,q(Yt), so that their topologies coincide. We can therefore say that P2(Yt) is compact in the weak topology defined as the topology of the dual space of Cq(Yt).
Paper not yet in RePEc: Add citation now
Ropponen, O. (2011), ‘Reconciling the evidence of card and krueger (1994) and neumark and wascher (2000)’, Journal of Applied Econometrics 26(6), 1051–1057.
- Santambrogio, F. (2015), ‘Optimal transport for applied mathematicians’, Progress in Nonlinear Differential Equations and their applications 87.
Paper not yet in RePEc: Add citation now
- Santambrogio, F. & Wang, X.-J. (2016), ‘Convexity of the support of the displacement interpolation: Counterexamples’, Applied Mathematics Letters 58, 152–158.
Paper not yet in RePEc: Add citation now
- Shapiro, A. (1991), ‘Asymptotic analysis of stochastic programs’, Annals of Operations Research 30(1), 169–186.
Paper not yet in RePEc: Add citation now
- Sinha, A., Malo, P. & Deb, K. (2017), ‘A review on bilevel optimization: From classical to evolutionary approaches and applications’, IEEE Transactions on Evolutionary Computation 22(2), 276–295.
Paper not yet in RePEc: Add citation now
- The function W2 2 (P, PYjt ) is continuous in P in the weak topology (with respect to Cq(Yt)) for fixed {PYjt }j=2,...,J+1. This follows from the fact that convergence in the Wasserstein metric on compact sets is equivalent to weak convergence (Santambrogio 2015, Theorem 5.9). Continuous functions on compact sets obtain their optimum, which shows that P(λ) exists for each λ ∈ ∆J−1 . Uniqueness of P(λ) for every λ ∈ ∆J−1 then follows from Proposition 3.5 in Agueh & Carlier (2011) under the assumption that at PYjt is absolutely continuous with respect to Lebesgue measure for j = 2, . . . , J + 1 and t ≤ T0.
Paper not yet in RePEc: Add citation now
- URL: https://github.com/rflamary/POT Folland, G. B. (2013), Real analysis: modern techniques and their applications, John Wiley & Sons.
Paper not yet in RePEc: Add citation now
- van der Vaart, A. & Wellner, J. (2013), Weak convergence and empirical processes: with applications to statistics, Springer Science & Business Media.
Paper not yet in RePEc: Add citation now
- Villani, C. (2003), Topics in Optimal Transportation, Graduate Studies in Mathematics vol. 58, American Mathematical Society.
Paper not yet in RePEc: Add citation now
- Villani, C. (2008), Optimal transport: old and new, Vol. 338, Springer Science & Business Media.
Paper not yet in RePEc: Add citation now
- Viviano, D. & Bradic, J. (2019), ‘Synthetic learner: model-free inference on treatments over time’, arXiv:1904.01490 .
Paper not yet in RePEc: Add citation now
- where C0(Rd ) is the space of all functions with vanishing tails. By a standard Rieszrepresentation (Aliprantis & Border 2006, chapter 14), one can identify dual space of Cq,0 by the space of all measures on Rd with finite second moments, which we define by M2(Rd ). P2(Rd ) is then defined as M2(Rd )∩P(Rd ), i.e. as the intersection of the dual space of Cq,0 with the positive cone of all probability measures on Rd . A.1 Proof of Proposition 1 We split the proof into two lemmas and the main proof: Lemmas 2 and 3 show the existence of a solution of (3) and (4) and the main proof shows the uniqueness of the counterfactual PN Y1t . Lemma 2. Under Assumptions 1 and 2, the minimum in min P∈P2(Rd) J+1 X j=2 λj W2 2 (P, PYjt ), t ≤ T0 is attained by a unique P(λ), which is continuous in λ ∈ ∆J−1 in the weak topology for the dual of Cq.
Paper not yet in RePEc: Add citation now
Zhou, H., Li, L. & Zhu, H. (2013), ‘Tensor regression with applications in neuroimaging data analysis’, Journal of the American Statistical Association 108(502), 540– 552.