Stochastic Frontier Model (SFA) in STATA

แชร์
ฝัง
  • เผยแพร่เมื่อ 7 ก.พ. 2025
  • This video demonstrates the true-fixed effects stochastic frontier model in STATA.

ความคิดเห็น • 111

  • @kwadwoarhin4822
    @kwadwoarhin4822 6 ปีที่แล้ว +3

    Very intuitive. Another video on other assumptions (gamma, truncated or half normal) on the inefficient term, u, and under what conditions each may be appropriately employed would be very much appreciated. Thank you

  • @kavitasarvesh8315
    @kavitasarvesh8315 4 ปีที่แล้ว +1

    Thank you for explaining the SFA in very simple terms. Look forward to more with other than TFE model.

  • @meenakshiparida1596
    @meenakshiparida1596 3 ปีที่แล้ว +1

    Thank you so much for this awesome explanation. I request you to kindly come up with a video explaining the input-specific technical efficiency estimation using the random coefficients frontier approach.

  • @davidayieko3864
    @davidayieko3864 6 ปีที่แล้ว +1

    Good work Brian.You explanations are well understood.Consider more videos.

  • @rkmofficial202
    @rkmofficial202 4 ปีที่แล้ว

    Nice explanation in simplest way and simple English

  • @biraraendalew
    @biraraendalew ปีที่แล้ว

    Wonderful video to understand and run technical efficiency.

  • @Abel_The_
    @Abel_The_ 5 ปีที่แล้ว +2

    really helpful brother keep doing your way of presentation is really interesting thank you so much

  • @SagangaKapaya
    @SagangaKapaya 2 ปีที่แล้ว +1

    Thank you. Very clear explanations, helped a lot.

  • @TÔMTIÊNYÊN
    @TÔMTIÊNYÊN ปีที่แล้ว +1

    thank you, clear

  • @mateenmateen8047
    @mateenmateen8047 4 ปีที่แล้ว +1

    You made my life much easier. God bless you bro.❤️❤️

  • @jairammeena4440
    @jairammeena4440 4 ปีที่แล้ว +1

    very clear video lecture even for the beginners.

  • @elwaleedahmaed6349
    @elwaleedahmaed6349 6 ปีที่แล้ว +1

    Thank you so much for the great explanation of sfa. God bless you

  • @theaurasian6708
    @theaurasian6708 6 ปีที่แล้ว +1

    Wonderful and very helpful presentation. Thanks so much!

  • @dr.ernestwinful1257
    @dr.ernestwinful1257 4 ปีที่แล้ว

    Very useful explanation. straight to the point

  • @rupanandawidanage947
    @rupanandawidanage947 6 ปีที่แล้ว +9

    Great explanation. Can you provide an example of cross sectional data. Thank you

  • @tawandachiwetu8752
    @tawandachiwetu8752 6 ปีที่แล้ว +2

    you sir are a genius. God bless you.

    • @brianmazorodze4860
      @brianmazorodze4860  5 ปีที่แล้ว +1

      Amen

    • @houdafaris4760
      @houdafaris4760 5 ปีที่แล้ว

      @@brianmazorodze4860 Can you Please show us how to apply "Data Envelopment Analysis (DEA)" in another video ?

  • @mannujain7604
    @mannujain7604 3 ปีที่แล้ว +1

    wonderful explanation

  • @mdc2scmobiledatacollection199
    @mdc2scmobiledatacollection199 2 ปีที่แล้ว

    Great, you are one of the best

  • @nnovationi3414
    @nnovationi3414 4 ปีที่แล้ว

    thank u so much for your vedios, sovled my problems!

  • @fleurydushime806
    @fleurydushime806 2 ปีที่แล้ว

    Wonderful presentation. My question is, is it possible to measure the technical efficiency for one single service? Thank you for answering

    • @brianmazorodze4860
      @brianmazorodze4860  ปีที่แล้ว

      One way is to simply estimate a semi-log model. The alternative is to transform your variable first by adding a constant that eliminates negative values before taking logs.

  • @nurbekomonov660
    @nurbekomonov660 ปีที่แล้ว

    Thank you! great explanation

  • @samuelwatson4624
    @samuelwatson4624 4 ปีที่แล้ว +2

    I appreciate for your awesome contribution! Could you please provide an example about semi-parametric frontier? Thank you!!!

  • @AnkitKumar-vf2hs
    @AnkitKumar-vf2hs 5 ปีที่แล้ว +1

    Thank you...
    Great explanation..

  • @dr.kashikaarora8454
    @dr.kashikaarora8454 4 ปีที่แล้ว +1

    Nicely explined video. Could you take interactive terms also and then explain SFA

  • @mercedesfrancesvaldiviajap6594
    @mercedesfrancesvaldiviajap6594 5 ปีที่แล้ว +1

    Thanks a lot !!! God bless you!!!

  • @SelemaniOmari
    @SelemaniOmari 2 ปีที่แล้ว

    Thank you Brian. Please explain how to use HAUSMAN TEST with sfpanel between TFE and TRE. Is there any simple way in Stata to produce result of Hausman Test?

  • @junnguyen1286
    @junnguyen1286 3 ปีที่แล้ว

    Thank you very much for your sharing. Could you please explain to me what is the difference in simulation of technical function and cost function? The video only shows the way to simulate the technical function. Thanks a lot for your help

  • @goodlifealways1737
    @goodlifealways1737 6 ปีที่แล้ว +1

    thank you so much for the great video. Very informative. I would appreciate if you helped with commands on cross sectional data just as you did with panel data.

  • @denfordmusekwa4175
    @denfordmusekwa4175 6 ปีที่แล้ว +1

    its awesome doc.. l appreciate

  • @amandakristabela9616
    @amandakristabela9616 4 ปีที่แล้ว +1

    Hi sir thank you for the clear explanation, btw sir, how to determine which one should we use between TRE and TFE?

    • @brianmazorodze4860
      @brianmazorodze4860  4 ปีที่แล้ว

      Use the normal Hausman test... the concept remains the same

  • @daipavandhar4802
    @daipavandhar4802 2 ปีที่แล้ว

    Also can I put more than one variable to measure technical inefficiency?

  • @idehengeorge1676
    @idehengeorge1676 3 ปีที่แล้ว

    Thank you very much.
    How do we compute the allocative efficiency?

  • @jyotsna5276
    @jyotsna5276 4 ปีที่แล้ว +1

    Thank u brian

  • @hemantdash7132
    @hemantdash7132 ปีที่แล้ว

    very useful..but can you please explain how to estimate input oriented technical efficiency using stata.. Is it only changing the orientation from output to inputs? How can solve translog production function? please explain with an example.

  • @mosh71
    @mosh71 5 ปีที่แล้ว +1

    Sir, if you are familiar, pls also make a video to explain how to do an input distance function. For example in the case of electricity, may be some power plants use more fuel to produce less electricity. In that case we need to increase fuel efficiency instead of increasing output. I am wondering that may not fit in this model, would it? (will appreciate any answer).

  • @daipavandhar4802
    @daipavandhar4802 2 ปีที่แล้ว

    Thank you for explaining this. Can I get robust standard errors for the results of stochastic frontier model? That is, can I use the commands "cluster(variable name)" and "robust" for panel data and cross section data respectively?

  • @basheerkkw
    @basheerkkw 3 ปีที่แล้ว

    thank you so much for your valuable video. why is it that stata generated for command "predict te" values greater than one? can you kindly please help me to solve this?

  • @nadeemahmadbhat5375
    @nadeemahmadbhat5375 6 ปีที่แล้ว +2

    Thank you sir
    How we can calculate Vi and thus full error term and what abt heteroscedaticity and exogenous determinants

    • @brianmazorodze4860
      @brianmazorodze4860  5 ปีที่แล้ว

      Thanks for your question. I will do a video on that, it is just an extension on the command by including the var specification in which het correlates are included as additional variables.

  • @harshadasanayake7378
    @harshadasanayake7378 5 ปีที่แล้ว +2

    Hi Brian, thanks for the wonderful video. I am using SFA to measure Tax efficiency. Could you pls explain me how to use the latest development in SFA by Kumbhakar, Lien and Hardaker (2014) which allows the separation between country effects, persistent inefficiency and time varying inefficiency. Thanks in advance.

    • @jnyaboe7215
      @jnyaboe7215 5 ปีที่แล้ว

      Hello Harsha, did you get a breakthrough with this?

  • @jyotsnarosario2204
    @jyotsnarosario2204 2 ปีที่แล้ว

    Can you please explain true random effects model? How is theta interpreted in tre models

  • @nigussietefera1077
    @nigussietefera1077 5 ปีที่แล้ว

    Thanks a lot for a very good explanation. It is clear

  • @ammarizohir7328
    @ammarizohir7328 6 ปีที่แล้ว +1

    thank you

  • @mircl3269
    @mircl3269 ปีที่แล้ว

    Hello, @Brian Mazorodze!!
    What is the command in stata to estimate efficiencies of cross-sectional data with the maximum likelihood estimator?

  • @martha_wainans
    @martha_wainans ปีที่แล้ว

    I wish you couyld share the dataset so that one can run the model practically

  • @karakesteven6617
    @karakesteven6617 4 ปีที่แล้ว +1

    I have liked this but how can we use the stochastic frontier model to estimate total factor productivity in the service sector with 10 sub-sectors in the entire sector? Thanks

  • @alpersinancalik6787
    @alpersinancalik6787 2 ปีที่แล้ว

    Any possibility of sharing the data so I can try it too

  • @alulaable
    @alulaable 6 ปีที่แล้ว +1

    great effort!

  • @SagangaKapaya
    @SagangaKapaya ปีที่แล้ว +1

    I have a small confusion, please help, the auxiliary equation for estimating inefficiency, is it not usigma(varlist_u[, noconstant]), instead of emean(varlist_m[, noconstant]) you are using, I was looking in the STATA documentation comparing to what you did here.

    • @brianmazorodze4860
      @brianmazorodze4860  ปีที่แล้ว

      Greetings. Firstly let me mention that [,noconstant] is optional as you probably already know. Similar to any standard regression, it is an option that lets you decide whether you want to estimate the inefficiency specification through the the origin or not. Then secondly, usigma(varlist.....) is relevant in cases where you see the need to model heteroscedasticity i.e., if you think that the z variables (or some of them) affect both the conditional mean (emean) and the variance of the inefficiency term (usigma). If one is to look at this example, i am implicitly assuming that the z variables do not affect the variance of the inefficiency term. However, it is important to note that there are some one-step specifications in which z can only affect inefficiency through the variance (i.e., usigma). Examples include models proposed by Mustafa U. Karakaplan. Maybe you saw a class of these models i am unsure without further clarification. Cheers.

    • @SagangaKapaya
      @SagangaKapaya ปีที่แล้ว +1

      @@brianmazorodze4860 Thanks again, I really appreciate for the video, reply and clarification. Very helpful.

  • @sibanideka8883
    @sibanideka8883 6 ปีที่แล้ว +1

    thank you sir. Can you provide another video regarding parametric malmquist productivity index

  • @joshymathewk2764
    @joshymathewk2764 ปีที่แล้ว

    Thank you for the video. But how to do SFA when there are -ve values in data and we cannot generate its log.

    • @brianmazorodze4860
      @brianmazorodze4860  ปีที่แล้ว

      You may have to consider adding a constant number that ensures you have positive values before taking logs.

    • @brianmazorodze4860
      @brianmazorodze4860  ปีที่แล้ว

      i.e, if you have Y log (Y+m) where m is a constant number that ensures that all values of Y are positives.

    • @joshymathewk2764
      @joshymathewk2764 ปีที่แล้ว

      Thank you sir. Will try that. Any possibility of contacting you sir, to clear doubts in SFA in stata?

  • @ishitasachdeva8287
    @ishitasachdeva8287 5 ปีที่แล้ว

    Hi, can you explain how to apply the mundlak transfromation of Baetts and Coelli model in stata ?

  • @charleslwanga3854
    @charleslwanga3854 5 ปีที่แล้ว

    Dear Brian, can i use this frontier analysis model to measure say managers efficiency in an organisation? How? Suppose i create the input variable to be what they are set out to do and score them out of 100% and also score the output out of 100% as an output variable, then have years of management experience as the Experience variable. Then, use cross sectional data and 'sf' command as you have explained. I would also like to capture whether the manager has basic skill in human resources as a variable. Give me some guide, thank you.

  • @mohammadhanif6589
    @mohammadhanif6589 4 ปีที่แล้ว

    Anyone know why when I use predict jlms command the numbers are the same? Im using model bc92 btw

  • @khabophilie7455
    @khabophilie7455 4 ปีที่แล้ว +2

    I'm doing my research on " technical efficiency of sugarcane farm irrigators" I have no idea on how to simplify the model to write my proposal....I wish you can help me

    • @brianmazorodze4860
      @brianmazorodze4860  4 ปีที่แล้ว +1

      I can help you. Write me an email. brianmazorodze@gmail.com

  • @andryarif4114
    @andryarif4114 ปีที่แล้ว

    Helo Brian thanks for the video,
    i have a question, what is the command to calculate the te? is it "Predict te, jlms ?"

    • @brianmazorodze4860
      @brianmazorodze4860  ปีที่แล้ว

      Yes, and you replace jlms with bc if you want efficiency scores produced based on battese and coeli

    • @andryarif4114
      @andryarif4114 ปีที่แล้ว

      @@brianmazorodze4860 the "Predict te, jlms" command is refered to Maximum Likelihood Estimation with Half Normal Distribution by Kumbhakar, S.C right?

  • @rupanandawidanage947
    @rupanandawidanage947 5 ปีที่แล้ว +1

    I have data on potato farming in Guatemala for 2016 and 2017 on 105 farmers. Can I conduct panel data (Stochastic frontier) for this data set?

    • @brianmazorodze4860
      @brianmazorodze4860  5 ปีที่แล้ว

      Yes you can depending on what transformation you will do, for T=2 N=105 as is the case here, a difference transformation will reduce your dataset to cross sectional. A TRE specification or TFE (within effect) might work better

    • @rupanandawidanage947
      @rupanandawidanage947 5 ปีที่แล้ว +1

      @@brianmazorodze4860 Thank you very much for responding. very appreciate your response.

  • @maambomwiinga8565
    @maambomwiinga8565 5 ปีที่แล้ว +1

    WELL UNDER STOOD. YOU SHOULD ALSO TALK ABOUT HOW TO PREDICT TECHNICAL EFFICIENCY

  • @arindambandyopadhyay3896
    @arindambandyopadhyay3896 4 หลายเดือนก่อน

    Where to get the .dta working file?

  • @houdafaris4760
    @houdafaris4760 5 ปีที่แล้ว +1

    Thank you, please can you show us how to apply "Data Envelopment Analysis (DEA)" in another video ?

  • @angelsharma5363
    @angelsharma5363 5 ปีที่แล้ว

    Can we have the dataset so we can follow what you are doing by doing it ourselves please?

  • @zeeshanali1799
    @zeeshanali1799 5 ปีที่แล้ว +1

    still waiting for translog function specification

  • @sindhubhutto5705
    @sindhubhutto5705 5 ปีที่แล้ว

    Sir kindly can u make video of Sfm model in gravity model

  • @zeeshanali1799
    @zeeshanali1799 5 ปีที่แล้ว +1

    Still waiting for translog function sir

  • @mazalale
    @mazalale 4 ปีที่แล้ว +1

    Any possibility of sharing the data so I can try it?

    • @brianmazorodze4860
      @brianmazorodze4860  4 ปีที่แล้ว

      Yes sure. Drop your email.

    • @SelemaniOmari
      @SelemaniOmari 4 ปีที่แล้ว

      @@brianmazorodze4860 please sample data of both sfcross and sfpanel to ssomari@gmail.com. Thank you Brian

  • @zeeshanali1799
    @zeeshanali1799 5 ปีที่แล้ว +1

    sir kindly conduct analysis by trans log function too

    • @brianmazorodze4860
      @brianmazorodze4860  5 ปีที่แล้ว

      No problem. I will do that soon

    • @sindhubhutto5705
      @sindhubhutto5705 5 ปีที่แล้ว

      Hey do u have any idea how to apply SFM in gravity model

  • @proking1572
    @proking1572 3 ปีที่แล้ว

    Can u plz share the data for testing

  • @tombwire6918
    @tombwire6918 3 ปีที่แล้ว

    Why is import share not transformed to ln import?

    • @brianmazorodze4860
      @brianmazorodze4860  3 ปีที่แล้ว

      Because it is a ratio, ln(imports/output) would mean lnimports - lnOutput which would not make sense intuitively

  • @ghoomopakistan4482
    @ghoomopakistan4482 5 ปีที่แล้ว +1

    Dear Mazorodze
    i am a PhD Scholar and working on Stochastic Frontier Analysis Translog function
    kindly help me out and make a video on translog function

  • @zeeshanali1799
    @zeeshanali1799 5 ปีที่แล้ว

    still waiting for translog function

  • @dollygaur2341
    @dollygaur2341 5 ปีที่แล้ว +3

    Hello sir!! Can u please provide ur email id? I need help with technical efficiency calculation for banking sector.....

  • @esterkeraru17
    @esterkeraru17 3 ปีที่แล้ว +1

    Thank you for this video, really helping me. May I consult with you via email, please? Thx

  • @rosettekhalil2586
    @rosettekhalil2586 4 ปีที่แล้ว +1

    A lot of noise around you !

    • @brianmazorodze4860
      @brianmazorodze4860  4 ปีที่แล้ว

      sorry about that

    • @rosettekhalil2586
      @rosettekhalil2586 4 ปีที่แล้ว

      Thanks for the video it’s quite interesting but here we are not able to to know how you linked between technical efficiency and external factors ? I hope you can give me an idea.. And what about translog function ! Thanks

  • @ratninja09
    @ratninja09 4 ปีที่แล้ว

    your name ain't brian