1
$\begingroup$

Since we need to consider 'Endogeneity' between price and quantity while calculating price elasticity and since linear regression cannot handle the phenomenon of endogeneity if objective of the model is to find causality, instrumental variable regression is used.

Since theory backs the idea of using IV regression, does this happens in real world? I mean corporations and governments use IV regression for price elasticity or do they use linear regression despite it's shortcomings in it's inablity to handle endogeneity?

$\endgroup$
  • 1
    $\begingroup$ I'm voting to close this question as off-topic because it has been cross-posted to another SE site: stats.stackexchange.com/questions/197838/…. Cross-posting is generally discouraged within the Stack Exchange network. $\endgroup$ – Ubiquitous Feb 24 '16 at 12:06
  • 1
    $\begingroup$ @Ubiquitous, i have followed the instructions in this meta discussion and made it suitable for the other forum instead of copy-pasting question word by word as suggested by Scortchi in the accepted answer. Thanks $\endgroup$ – Enthusiast Feb 25 '16 at 10:21
1
$\begingroup$

They use all types of methods.

Often, papers will use a very basic linear regression model and submit to a low prestige journal so one can justify ones grant and boost the H index as much as possible when meta analyses cite your article to clean up the mess.

It's up to the meta analysis to take into account that not all models have correct methodology and create a meta-analytical model to correct for all these inaccuracies.

Fortunately, there are sometimes dynamic lag models and the like to compare to and guess at an appropriate correction term for the others.

Here is an example for gas prices.

https://www.google.com/url?sa=t&source=web&rct=j&url=http://papers.tinbergen.nl/06106.pdf&ved=0ahUKEwia2Me-85LLAhUUzmMKHWvjAd0QFgghMAE&usg=AFQjCNHi6RH_xX7h9NLPcVRCFAMP0OtERQ&sig2=3ChFSPGwukN_WC4IlhitGA

Edit: A meta analytic model is a model to explain why studies obtained the results that they did. Different samples, methodologies, and so forth will produce different results, and a metamodel expresses this formally.

A dynamic lag model (also called a distributed lag model) is a model to predict the current value of a dependent variable based on past values of independent variables. For example, if a change in price causes a change in demand, you can model this by correlating price at time t-1 to demand at time t.

$\endgroup$
  • 1
    $\begingroup$ Can you please elaborate on what is 1) Meta-analytic model and 2) dynamic lag model for calculating price elasticity and cross-price elasticity? $\endgroup$ – Enthusiast Feb 28 '16 at 3:44
  • $\begingroup$ Thanks! I was waiting for your comment and thought i will get notification someday. Didn't knew that you have already mentioned your answer by editing the post. I will mark this as complete now. Thanks for your help!! :-) $\endgroup$ – Enthusiast Mar 13 '16 at 10:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.