@article{fdi:010076434,
  title = {{E}stimating microcredit impact with low take-up, contamination and inconsistent data : a replication study of {C}r{\'e}pon, {D}evoto, {D}uflo, and {P}ariente ({A}merican {E}conomic {J}ournal: {A}pplied {E}conomics, 2015)},
  author = {{B}{\'e}d{\'e}carrats, {F}. and {G}u{\'e}rin, {I}sabelle and {M}orvan-{R}oux, {S}. and {R}oubaud, {F}ran{\c{c}}ois},
  editor = {},
  language = {{ENG}},
  abstract = {{W}e replicate a flagship randomised control trial carried out in rural {M}orocco that showed substan- tial and significant impacts of microcredit on the assets, the outputs, the expenses and the profits of self-employment activities. {T}he original results rely primarily on trimming, which is the exclusion of observation with the highest values on some variables. {H}owever, the applied trimming proce- dures are inconsistent between the baseline and the endline. {U}sing identical specifications as the original paper reveals large and significant imbalances at the baseline and, at the endline, impacts on implausible outcomes, like household head gender, language or education. {T}his calls into ques- tion the reliability of the data and the integrity of the experiment protocol. {W}e find a series of coding, measurement and sampling errors. {C}orrecting the identified errors lead to different results. {A}fter rectifying identified errors, we still find substantial imbalances at baseline and implausible impacts at the endline. {O}ur re-analysis focused on the lack of internal validity of this experiment, but several of the identified issues also raise concerns about its external validity.},
  keywords = {{MAROC}},
  booktitle = {},
  journal = {{I}nternational {J}ournal for {R}e-{V}iews in {E}mpirical {E}conomics - {IREE}},
  volume = {3},
  numero = {3},
  pages = {53  [en ligne]},
  ISSN = {2566-8269},
  year = {2019},
  DOI = {10.18718/81781.12},
  URL = {https://www.documentation.ird.fr/hor/fdi:010076434},
}