Drawing the map…
Drawing the map…
Fetching the papers and their links…
96 papers grouped into meanings · 23 typed relationships (19 verified, 4 inferred)
Every paper and every typed relationship, with the evidence behind each. For the readable account, go back to the map. This page is long on purpose: nothing is hidden, so your browser’s find will search every quote on it. To go straight to one paper, jump to a paper.
Ranked by how often later work cites each paper for this term, not by how often it is cited overall. The first list is what the field built on. The second is where the meaning changed, and those papers are usually missing from the first list, because a paper that moves a term gets cited by the field it moved into, not by the one it left.
Provides an econometrics perspective on instrumental variables that connects the older, economics-based simultaneous equations literature (supply and demand models) to modern statistical applications such as randomized experiments with noncompliance.
For the first time in this class of problems, we derive optimal convergence rates, and show that they are attained by particular estimators.
Uses invariance arguments to construct a new minimum distance objective function for many-instrument IV inference
The paper extends prior identification results to identify conditional average potential outcomes using a dual pair of conditional moment restrictions (treatment bridge and outcome bridge), one of which is described as 'fully original.'
We propose an alternative solution to this problem, which we term Instrumental Variable Bayesian Model Averaging (IVBMA).
Derives novel confidence intervals/standard error formulas for shift-share instrumental variables estimators that are valid under arbitrary cross-regional correlation in residuals, extending the OLS results to the IV case where the shift-share variable is used as an instrument.
First paper to establish root-n consistency and asymptotic normality of a Lasso/Post-Lasso-based IV estimator for a low-dimensional structural parameter in a high-dimensional setting without imposing the restrictive 'beta-min' condition.
Defines weak identification for linear IV models with many instruments via necessary and sufficient conditions for existence of a consistent test based on the concentration parameter divided by the square root of the number of instruments.
Each row is a typed relationship between two papers. Click a row to see the evidence.
The same three marks stand beside each paper’s own passage, where they are about that passage alone: a passage marked ● can sit above a relationship marked ○.
“Suppose, however, that for each i we have available another observed data value, Wi, say (an instrumental variable), for which E(Ui|Wi) = 0 (1.2) and there is a “sufficiently strong” relationship between Xi and Wi.”
“If V is an instrument satisfying E[ Y( t) |X,V] =E [ Y( t) |X], then for any fixed x, bounds on θ^∗:= θ^∗(x):= E[ Y( t) |X=x] are given by sup_v∈𝒱θ ^l(x,v) ≤θ^∗(x) ≤inf_v∈𝒱θ ^u(x,v),”
“The model is y_i = d_i'α_0 + ϵ_i where α_0 denotes the true value of a vector-valued parameter α. y_i is the response variable, and d_i is a finite k_d-vector of variables whose first k_e elements contain endogenous variables. The disturbance ϵ_i obeys for all i (and n): E[ϵ_i|x_i] = 0, where x_i is a k_x-vector of instrumental variables.”
“and W are instrumental variables.”
“The second one is that the composite error U_it in the observed variable equation […] is correlated with the lagged dependent variable Y_it-1 and we may therefore need to use instrumental variables (IVs).”
“We consider the problem of incorporating instrument and covariate uncertainty into the Bayesian estimation of an instrumental variable (IV) regression system.”
“Second, the covariates may, in general, include instruments for treatment status, but they are not known as such. This is standard, but left implicit, in discussions of ignorability. If instruments are present, and selected for estimation, efficiency suffers but unbiasedness is not harmed. Efficiency bounds in this context typically (implicitly) assume there are no instruments in X.”
“Instrumental Variables (IV) refers to a set of methods developed in econonometrics starting in the 1920s to draw causal inferences in settings where the treatment of interest cannot be credibly viewed as randomly assigned, even after conditioning on additional covariates, that is, settings where the assumption of no unmeasured confounders does not hold.”
“I analyze a linear instrumental variables model with a single endogenous regressor and many instruments.”
“In this section, we employ the instrumental variable quantile regression (IVQR) method for estimation. Let d_it=∑_j≠ im_ijy_jt denote a scalar endogenous variable, which is related to a vector of instruments ω_it. The instruments ω_it are independent of ε_it.”
“There exists a vector Z of observable random variables, such that (Y_0,Y_1)⊥⊥ Z. Such variables are akin to typical instrumental variables, and examples within Roy models in the existing literature include parental education in […], distance to a college in […] and attendance in a Catholic high school in […].”
“which is the IV estimand using $h(Z)$ as instruments for $X.$”
“In particular, V and Z must each satisfy an instrumental relevance condition in an IV model in which W is a vector of endogenous regressors, with X and D acting as exogenous regressors.”
“Many studies use shift-share (or “Bartik”) instruments, which average a set of shocks with exposure share weights. We provide a new econometric framework for shift-share instrumental variable (SSIV) regressions in which identification follows from the quasi-random assignment of shocks, while exposure shares are allowed to be endogenous.”
“We also derive an analogous formula when X_i is used as an instrument in an instrumental variables regression, which follows directly from the fact that the associated first-stage and reduced-form regressions take the form in […].”
“as long as Z_t satisfies conventional assumptions of […], we can establish a causal link between Y_it and W_it by constructing an instrumental variables (IV) estimator separately for each unit i and reporting a summary of these estimators, e.g., the average.”
“Several authors have thus tried to use instrumental variable approaches to estimate the effect of CBI on inflation within a causal framework, but have been unable to find strong instruments […].”
“We show that the high-dimensional slope parameter of a high-frequency covariate can be identified and accurately estimated leveraging on a low-frequency instrumental variable.”
“They first remove factor loadings from the estimation equation and then estimate the remaining common factors and parameters using lagged regressors as instruments.”
“We develop a concept of weak identification in linear IV models in which the number of instruments can grow at the same rate or slower than the sample size.”
“We also assume the availability of valid external IVs (proxy variables) – variables that correlate with the shock of interest, but not with the other shocks.”
“Although this moment equation looks very similar to moment equations in instrumental variable estimation […], it is based on substantially different assumptions. Actually, post-treatment outcomes Y_𝚙𝚘𝚜𝚝 must not be valid instrumental variables, since below they are assumed to be strongly dependent with the unmeasured confounders.”
“In addition, we also present a couple of simpler examples with the linear regression models and the linear IV regression models in Appendix […].”
“In virtually all applications, the relation is assumed to contain an additive error term that is shown (e.g., by the assumption of Rational Expectations (RE)) or primitively assumed to be uncorrelated with predetermined variables excluded from the specified relation. This makes any predetermined variable a valid IV.”
“As expected, the median bias for SVAR-IV is particularly elevated relative to other estimation methods if the degree of invertibility is small, as predicted by theory.”
“A valid instrumental variable (IV) ought to have two essential qualities: first, it shows a high correlation with the corresponding endogenous explanatory variable; second, it fulfils the exclusion restriction”
“with j = { 1,2 }, and where W is a vector of instrumental variables.”
“The instrumental variable (IV) method is widely used when the treatment variable of interest is endogenous. As shown in Figure […], the ideal IV needs to be correlated with the endogenous treatment variable (C1), it should not have a direct effect on the outcome (C2) and should not be related to unobserved confounders that affect both outcome and treatment (C3).”
“even if D is endogenous i.e., Cov[D,ε]≠ 0, estimating λ by 2SLS using an instrument Z that is uncorrelated with ε does not guarantee a causal interpretation.”
“where the instrumental variable Z_i, exc is required to be exogenous with respect to ϵ_i, relevant for X_i, and excluded from the regression model […].”
“For general dynamic panel models, the Nickell bias is usually addressed with the instrumental variable (IV) method […] or analytical formulas […]. The IV method is subject to the lack of efficiency and performs poorly with finite samples when the regressor is persistent […].”
The newer paper generalizes the classical IV relevance/exogeneity conditions to proxy controls for unobserved confounders, building on the earlier nonparametric IV framework.— the tool’s reading
“Suppose, however, that for each i we have available another observed data value, Wi, say (an instrumental variable), for which E(Ui|Wi) = 0 (1.2) and there is a “sufficiently strong” relationship between Xi and Wi.”
“In particular, V and Z must each satisfy an instrumental relevance condition in an IV model in which W is a vector of endogenous regressors, with X and D acting as exogenous regressors.”
The newer paper generalizes the earlier exogeneity-based instrument concept into a broader conditional moment restriction framework with rank/completeness conditions for identification.— the tool’s reading
“Identification and estimation of linear nonparametric conditional moment models have been studied by Newey and Powell (2003), Hall and Horowitz (2005), Blundell, Chen, and Kristensen (2007), Darolles, Fan, Florens, and Renault (2011), and others.”
“Suppose, however, that for each i we have available another observed data value, Wi, say (an instrumental variable), for which E(Ui|Wi) = 0 (1.2) and there is a “sufficiently strong” relationship between Xi and Wi.”
“and W are instrumental variables.”
The newer paper restricts the general nonparametric exogeneity condition to a linear IV moment condition framework for identifying a specific linear approximation parameter.— the tool’s reading
“Suppose, however, that for each i we have available another observed data value, Wi, say (an instrumental variable), for which E(Ui|Wi) = 0 (1.2) and there is a “sufficiently strong” relationship between Xi and Wi.”
“which is the IV estimand using $h(Z)$ as instruments for $X.$”
The newer paper generalizes the earlier IV exogeneity condition into a broader linear inverse problem framework, embedding it as a leading example among other models.— the tool’s reading
“A leading example of this setting are nonparametric instrumental regressions [16, 17, 18], as mentioned above, but many other models, such as linear and non-linear parametric regressions and additive nonparametric regressions can fit in this general framework [?].”
“This framework can be easily extended to the case when S is an endogenous multivariate categorical variable and to include additional exogenous components in Z [17, 25, 26].”
“Suppose, however, that for each i we have available another observed data value, Wi, say (an instrumental variable), for which E(Ui|Wi) = 0 (1.2) and there is a “sufficiently strong” relationship between Xi and Wi.”
“with j = { 1,2 }, and where W is a vector of instrumental variables.”
The instrumental variable exclusion-restriction concept is carried from treatment-effect bounds into structural macro shock identification via proxy variables, a new modeling field.— the tool’s reading
“If V is an instrument satisfying E[ Y( t) |X,V] =E [ Y( t) |X], then for any fixed x, bounds on θ^∗:= θ^∗(x):= E[ Y( t) |X=x] are given by sup_v∈𝒱θ ^l(x,v) ≤θ^∗(x) ≤inf_v∈𝒱θ ^u(x,v),”
“We also assume the availability of valid external IVs (proxy variables) – variables that correlate with the shock of interest, but not with the other shocks.”
The newer paper generalizes the earlier exclusion-restriction instrument concept into a weaker stochastic monotonicity condition while reusing its intersection-bounds inference machinery.4— the tool’s reading
“(14) on the interquartile range in the randomized economy, is carried out with the STATA package clrbounds implementing Chernozhukov, Lee, and Rosen (2013).”
“It allows the incorporation of exclusion restrictions and lends itself to the partial identified inference of Chernozhukov, Lee, and Rosen (2013) and Andrews and Shi (2014).”
“As above, the bounds of proposition 5 are intersection bounds, so that inference can be carried out with the method proposed in Chernozhukov, Lee, and Rosen (2013).”
“If V is an instrument satisfying E[ Y( t) |X,V] =E [ Y( t) |X], then for any fixed x, bounds on θ^∗:= θ^∗(x):= E[ Y( t) |X=x] are given by sup_v∈𝒱θ ^l(x,v) ≤θ^∗(x) ≤inf_v∈𝒱θ ^u(x,v),”
“There exists a vector Z of observable random variables, such that (Y_0,Y_1)⊥⊥ Z. Such variables are akin to typical instrumental variables, and examples within Roy models in the existing literature include parental education in <cit.>, distance to a college in <cit.> and attendance in a Catholic high school in <cit.>.”
The newer paper carries the instrumental variable notion into the selection-on-observables/covariate framework, a different methodological context than IV estimation with many instruments.— the tool’s reading
“The model is y_i = d_i'α_0 + ϵ_i where α_0 denotes the true value of a vector-valued parameter α. y_i is the response variable, and d_i is a finite k_d-vector of variables whose first k_e elements contain endogenous variables. The disturbance ϵ_i obeys for all i (and n): E[ϵ_i|x_i] = 0, where x_i is a k_x-vector of instrumental variables.”
“Second, the covariates may, in general, include instruments for treatment status, but they are not known as such. This is standard, but left implicit, in discussions of ignorability. If instruments are present, and selected for estimation, efficiency suffers but unbiasedness is not harmed. Efficiency bounds in this context typically (implicitly) assume there…”
The newer paper carries the many-IV/weak-IV inference approach from cross-sectional Lasso-based instruments into time-series GMM settings with predetermined macroeconomic instruments like the NKPC.— the tool’s reading
“…contributes to the (very) many weak IVs literature predominantly restricted to the cross-sectional case (see Anatolyev and Gospodinov (2010), Belloni et al. (2012), Crudu, Mellace, and Sándor (2020), and Mikusheva and Sun (2020)), and can find application well beyond the example of NKPCs…”
“I now state the main theoretical result of this paper, which ensures that the approach proposed controls the size of the test.8 8It should be noted, however, that–similarly to other sup-based test statistics, such as in Belloni et al. (2012) and Theorem 1.”
“(see Belloni et al. (2012), Hansen and Kozbur (2014), and Ng and Bai (2009)).”
“The model is y_i = d_i'α_0 + ϵ_i where α_0 denotes the true value of a vector-valued parameter α. y_i is the response variable, and d_i is a finite k_d-vector of variables whose first k_e elements contain endogenous variables. The disturbance ϵ_i obeys for all i (and n): E[ϵ_i|x_i] = 0, where x_i is a k_x-vector of instrumental variables.”
“In virtually all applications, the relation is assumed to contain an additive error term that is shown (e.g., by the assumption of Rational Expectations (RE)) or primitively assumed to be uncorrelated with predetermined variables excluded from the specified relation. This makes any predetermined variable a valid IV.”
The newer paper describes a general quasi-differencing GMM approach with lagged regressors as instruments, of which the earlier paper's specific measurement-error correction is treated as a special application.— the tool’s reading
“Lee, Moon, and Weidner (2012) use the same estimator to account for measurement errors in the dependent variable in dynamic interactive fixed effects models.”
“Under assumptions similar to those in Moon and Weidner (2017), the authors show consistency and derive the asymptotic distribution of the minimum distance estimator.”
“The second one is that the composite error U_it in the observed variable equation <ref> is correlated with the lagged dependent variable Y_it-1 and we may therefore need to use instrumental variables (IVs).”
“They first remove factor loadings from the estimation equation and then estimate the remaining common factors and parameters using lagged regressors as instruments.”
The IV concept moves from dynamic panel endogeneity correction to spatial quantile regression, applying instruments to a spatial-lag regressor in a new modeling context.— the tool’s reading
“The second one is that the composite error U_it in the observed variable equation <ref> is correlated with the lagged dependent variable Y_it-1 and we may therefore need to use instrumental variables (IVs).”
“In this section, we employ the instrumental variable quantile regression (IVQR) method for estimation. Let d_it=∑_j≠ im_ijy_jt denote a scalar endogenous variable, which is related to a vector of instruments ω_it. The instruments ω_it are independent of ε_it.”
The newer paper applies the same IVBMA framework to a specific case using lagged endogenous regressors as instruments for democracy-related reverse causality, restricting the general method to a particular application domain.— the tool’s reading
“Among economic disciplines, growth researchers are pioneers in exploiting the capabilities of various Bayesian Model Averaging (BMA) techniques, including the IVBMA (Salai-Martin et al., 2004; Karl and Lenkoski, 2012).”
“Karl and Lenkoski (2012) incorporate uncertainty into both equations (7) and (8).”
“Two examples are Leon-Gonzalez and Montolio (2015) and Leon-Gonzalez and Vinayagathasan (2015). proposed by Karl and Lenkoski (2012).”
“We consider the problem of incorporating instrument and covariate uncertainty into the Bayesian estimation of an instrumental variable (IV) regression system.”
“A valid instrumental variable (IV) ought to have two essential qualities: first, it shows a high correlation with the corresponding endogenous explanatory variable; second, it fulfils the exclusion restriction”
The newer paper treats linear IV as a simple special case within a broader moment-restriction framework, restricting the earlier general instrumental-covariate discussion to a specific model example.— the tool’s reading
“(xxii) comrelig: Religious proximity (Disdier and Mayer, 2007) is an index calculated by adding the products of the shares of Catholics, Protestants and Muslims in the exporting and importing countries. It is bounded between 0 and 1, and is maximum if the country pair has a religion which (1) comprises a vast majority of the population, and (2) is the same i…”
“Second, the covariates may, in general, include instruments for treatment status, but they are not known as such. This is standard, but left implicit, in discussions of ignorability. If instruments are present, and selected for estimation, efficiency suffers but unbiasedness is not harmed. Efficiency bounds in this context typically (implicitly) assume there…”
“In addition, we also present a couple of simpler examples with the linear regression models and the linear IV regression models in Appendix <ref>.”
The newer paper applies the instrumental variable concept to continuous-time high-frequency econometric identification, a new domain, rather than directly extending the earlier covariate-efficiency discussion.ed— the tool’s reading
“Condition (iii) can also be relaxed to a weaker condition that the product of estimation errors converges at the oP(T−1/2) rate, in which case, one of estimators can be consistent at a slower rate provided that the other can offset it; see also Farrell (2015) and Chernozhukov et al. (2018).”
“Second, the covariates may, in general, include instruments for treatment status, but they are not known as such. This is standard, but left implicit, in discussions of ignorability. If instruments are present, and selected for estimation, efficiency suffers but unbiasedness is not harmed. Efficiency bounds in this context typically (implicitly) assume there…”
“We show that the high-dimensional slope parameter of a high-frequency covariate can be identified and accurately estimated leveraging on a low-frequency instrumental variable.”
The newer paper explicitly dispenses with the exclusion restriction central to standard IV, proposing an alternative identification strategy without excluded instruments.— the tool’s reading
“See, e.g., Imbens (2014), Imbens and Rubin (2015), Mogstad and Torgovitsky (2018), and Abadie and Cattaneo (2018) for more comprehensive reviews.”
“Imbens, G. (2014): “Instrumental variables: an econometrician’s perspective,” Tech. rep., National Bureau of Economic Research. Imbens, G. W. and J. D. Angrist (1994): “Identification and estimation of local average treatment effects,” Econometrica: journal of the Econometric Society, 467–475. Imbens, G. W. and D. B. Rubin (2015): Causal inference in statist…”
“Imbens, G. (2014): “Instrumental variables: an econometrician’s perspective,” Tech. rep., National Bureau of Economic Research. Imbens, G. W. and J. D. Angrist (1994): “Identification and estimation of local average treatment effects,” Econometrica: journal of the Econometric Society, 467–475. Imbens, G. W. and D. B. Rubin (2015): Causal inference in statist…”
“Instrumental Variables (IV) refers to a set of methods developed in econonometrics starting in the 1920s to draw causal inferences in settings where the treatment of interest cannot be credibly viewed as randomly assigned, even after conditioning on additional covariates, that is, settings where the assumption of no unmeasured confounders does not hold.”
“where the instrumental variable Z_i, exc is required to be exogenous with respect to ϵ_i, relevant for X_i, and excluded from the regression model <ref>.”
○ The check ran against the text this system pulled out of the papers and could not find one of these quotes, so this relationship is shown as inferred, never as verified.
The newer paper focuses specifically on validity conditions and failure cases of IV within linear models, restricting the broader IV identification framework to a specific analytic scenario.— the tool’s reading
“Instrumental Variables (IV) refers to a set of methods developed in econonometrics starting in the 1920s to draw causal inferences in settings where the treatment of interest cannot be credibly viewed as randomly assigned, even after conditioning on additional covariates, that is, settings where the assumption of no unmeasured confounders does not hold.”
“even if D is endogenous i.e., Cov[D,ε]≠ 0, estimating λ by 2SLS using an instrument Z that is uncorrelated with ε does not guarantee a causal interpretation.”
○ The check ran against the text this system pulled out of the papers and could not find one of these quotes, so this relationship is shown as inferred, never as verified.
The newer paper applies general IV identification theory to a specific case—lagged endogenous regressors within an IVBMA framework—restricting the general concept to a particular instrument choice and application domain.— the tool’s reading
“We refer readers to Angrist et al. (1996), Imbens (2014), and Athey and Imbens (2017) for thorough discussions about causal inference and instrumental variables methodology.”
“Instrumental Variables (IV) refers to a set of methods developed in econonometrics starting in the 1920s to draw causal inferences in settings where the treatment of interest cannot be credibly viewed as randomly assigned, even after conditioning on additional covariates, that is, settings where the assumption of no unmeasured confounders does not hold.”
“A valid instrumental variable (IV) ought to have two essential qualities: first, it shows a high correlation with the corresponding endogenous explanatory variable; second, it fulfils the exclusion restriction”
○ The check ran against the text this system pulled out of the papers and could not find one of these quotes, so this relationship is shown as inferred, never as verified.
The newer paper restricts the general IV identification framework to a specific empirical application (central bank independence and inflation) without extending or contesting the underlying method.— the tool’s reading
“y, starting with path analyses and potential outcome language (Tinbergen, 1930; Wright, 1934) and continuing with regression discontinuity analyses (Hahn et al., 2001), instrumental variable designs (Imbens, 2014), propensity score approaches in the context of the potential outcome framework (Rosenbaum and Rubin, 1983), among many other methods. More recentl…”
“Instrumental Variables (IV) refers to a set of methods developed in econonometrics starting in the 1920s to draw causal inferences in settings where the treatment of interest cannot be credibly viewed as randomly assigned, even after conditioning on additional covariates, that is, settings where the assumption of no unmeasured confounders does not hold.”
“Several authors have thus tried to use instrumental variable approaches to estimate the effect of CBI on inflation within a causal framework, but have been unable to find strong instruments <cit.>.”
○ The check ran against the text this system pulled out of the papers and could not find one of these quotes, so this relationship is shown as inferred, never as verified.
The newer paper builds on the linear IV framework by generalizing it to handle invalid/weak instruments and extends the MCD test from Kolesár (2018) for high-dimensional settings.— the tool’s reading
“According to Kolesár (2018) Proposition 4, the MCD test with asymptotic size %n Algorithm 1 WIT estimator with MCD test tuning strategy Input: Y ,Z,D,λseq, %n, and J 1: Calculate β̄ in (34), initialize α̂MCP = 1 and I = 1 2: for α̂(0) = 0, α̂(0)(β̌j) (35) do . β̌j ∈ β̄ in priority of the largest…”
“Kolesár et al. (2015); Kolesár (2018) allowed the number of covariates to grow with the sample size.”
“Therefore, we propose the modified Cragg-Donald (MCD, Kolesár, 2018) test-based tuning procedure, which extends the Sargan test to allow high-dimensional covariates and IVs.”
“I analyze a linear instrumental variables model with a single endogenous regressor and many instruments.”
“The instrumental variable (IV) method is widely used when the treatment variable of interest is endogenous. As shown in Figure <ref>, the ideal IV needs to be correlated with the endogenous treatment variable (C1), it should not have a direct effect on the outcome (C2) and should not be related to unobserved confounders that affect both outcome and treatment…”
The newer paper explicitly contrasts its bridge function moment conditions with instrumental variable assumptions, arguing its post-treatment variables violate IV exogeneity/relevance conditions.— the tool’s reading
“This concept is originally proposed in the negative control literature [Miao and Tchetgen, 2018, Cui et al., 2020, Tchetgen et al., 2020, Deaner, 2021, Miao et al., 2016], where some also apply it to panel data.”
“…5, we extend our methodology to alternative causal effect estimands and linear factor models with time-varying unmeasured factors, and connect this paper to the negative control framework proposed in recent literature [e.g., Cui et al., 2020, Tchetgen et al., 2020, Miao et al., 2018, Deaner, 2021].”
“Recently, a series of works propose a negative control framework to deal with the challenge of unmeasured confounding [e.g., Cui et al., 2020, Tchetgen et al., 2020, Miao et al., 2018, Deaner, 2021, Shi et al., 2020].”
“In particular, V and Z must each satisfy an instrumental relevance condition in an IV model in which W is a vector of endogenous regressors, with X and D acting as exogenous regressors.”
“Although this moment equation looks very similar to moment equations in instrumental variable estimation <cit.>, it is based on substantially different assumptions. Actually, post-treatment outcomes Y_𝚙𝚘𝚜𝚝 must not be valid instrumental variables, since below they are assumed to be strongly dependent with the unmeasured confounders.”
The newer paper generalizes the shift-share instrument framework to time-series instruments interacted with exposure measures within a broader 2SLS causal identification setting.— the tool’s reading
“Identification is now achieved exploiting variation over industries (see Borusyak et al. [2018]).”
“As a practical recommendation, we suggest that applied researchers use the robust t-test procedure developed in Ibragimov and Müller [2010], or the permutation-based test developed in Canay et al. [2017]. The standard approach to weak inference in the literature assumes that the first-stage and reduced-form estimators have a joint normal distribution with kn…”
“Many studies use shift-share (or “Bartik”) instruments, which average a set of shocks with exposure share weights. We provide a new econometric framework for shift-share instrumental variable (SSIV) regressions in which identification follows from the quasi-random assignment of shocks, while exposure shares are allowed to be endogenous.”
“as long as Z_t satisfies conventional assumptions of <cit.>, we can establish a causal link between Y_it and W_it by constructing an instrumental variables (IV) estimator separately for each unit i and reporting a summary of these estimators, e.g., the average.”
The newer paper generalizes the shift-share IV idea to interacted aggregate time-series instruments with unit exposures, broadening the classical shift-share framework.— the tool’s reading
“, Card and Krueger [1993], Abadie and Gardeazabal [2003], Abadie et al.”
“We also derive an analogous formula when X_i is used as an instrument in an instrumental variables regression, which follows directly from the fact that the associated first-stage and reduced-form regressions take the form in <ref>.”
“as long as Z_t satisfies conventional assumptions of <cit.>, we can establish a causal link between Y_it and W_it by constructing an instrumental variables (IV) estimator separately for each unit i and reporting a summary of these estimators, e.g., the average.”
Applies high-dimensional IV/GMM instrument framework from generic econometrics to macroeconomic rational-expectations models like NKPC, a new domain application...— the tool’s reading
“We develop a concept of weak identification in linear IV models in which the number of instruments can grow at the same rate or slower than the sample size.”
“In virtually all applications, the relation is assumed to contain an additive error term that is shown (e.g., by the assumption of Rational Expectations (RE)) or primitively assumed to be uncorrelated with predetermined variables excluded from the specified relation. This makes any predetermined variable a valid IV.”
The newer paper restricts the general IV/shock-identification concept to lagged internal instruments used specifically for bias correction in dynamic panel models, a special-case application.— the tool’s reading
“LP and its extensions have shaped a growing time series literature (Barnichon and Brownlees, 2019; Jordà, 2009; Xu, 2022), with systematic comparisons between LP and VAR theoretically (Montiel Olea and Plagborg-Møller, 2021; Plagborg-Møller and Wolf, 2021) and numerically (Li et al., 2022).”
“As expected, the median bias for SVAR-IV is particularly elevated relative to other estimation methods if the degree of invertibility is small, as predicted by theory.”
“For general dynamic panel models, the Nickell bias is usually addressed with the instrumental variable (IV) method <cit.> or analytical formulas <cit.>. The IV method is subject to the lack of efficiency and performs poorly with finite samples when the regressor is persistent <cit.>.”