3
$\begingroup$

I am reading numerous articles pertaining to unemployment as references for my own work. Yet I've encountered many where they use long time series in countries which have had some sort of pertinent regulatory adjustment. By that, I mean that there are laws and regulations that were enacted during the period of analysis. Yet, the time series are not broken into subdivisions to be studied independently.

$\endgroup$
  • $\begingroup$ Prime facie everything is 'inherently different'. But if you spell out what that phrase means to you, it might be easier to see what implicit theory you have in mind and thereby turn this into an answerable question. $\endgroup$ – conjugateprior Oct 8 '14 at 8:58
  • $\begingroup$ @conjugateprior I will defer to Lucas: "Given that the structure of an econometric model consists of optimal decision rules of economic agents, and that optimal decision rules vary systematically with changes in the structure of series relevant to the decision maker, it follows that any change in policy will systematically alter the structure of econometric models." $\endgroup$ – erasmortg Oct 8 '14 at 9:23
  • $\begingroup$ Defer as you like. Even if you accept the premises of that sentence - most would accept that the first is descriptively incorrect about actual economic agents (the second seems true by definition) - it still remains to unpack the idea of 'systematically alter'. You seem to want to read it as implying 'we should break up time series and analyse the parts separately'. That does not seem to follow from the quote, though it might be reasonable on some other background theory. I was suggesting you say what that theory was. $\endgroup$ – conjugateprior Oct 8 '14 at 9:36
  • $\begingroup$ Grammar aside, I don't want to tinker with the title, but it does not seem what you are reaching towards, which is more perhaps like why are analyses based on time series considered valid when the underlying phenomena change? A short answer is: Good question, and many researchers are dubious on how far studying time series by themselves can deliver much quantitative insight, and they correspondingly prefer to build models that include time-varying predictors. But you should at the outset look at what you are trying to explain. $\endgroup$ – Nick Cox Oct 8 '14 at 10:19
3
$\begingroup$

The world is changing constantly, so if you want to apply time series analysis only when the environment is constant you cannot do time series analysis at all. This does not mean that you can ignore changes in the environment, but it does mean that it is necessarily a trade-off when to split a time-series up. Apperently, these authors made a different trade-off than you would have done. That happens all the time. You can discuss that in your literature review. But I would only do that if it helps bring across the message you want to convey in your article, otherwise it is just a distraction.

$\endgroup$
3
$\begingroup$

If what one writes is (socio)economic history, and wants to accompany the narration and analysis with summary quantitative, descriptive measures, "breaking the time series in subdivisions in order to be studied", appears helpful and sound -because here, the model used to extract these measures from the data is not the prime explanatory force, or focus, of the research endeavor.

But if one tries to compact the analysis and the arguments inside an economic/econometric model, then this "breaking" inescapably denies that, alongside things that change, there are other things that remain the same, as an economy evolves.

This is why there are numerous econometric techniques that have been developed in order to capture phenomena like the one you describe: from the inclusion of dummy variables in the specification to "take on" the effect of the special temporary phenomenon, to "structural breaks" models (with maybe multiple breaks), to "regime-switching" models, all try to study in a comprehensive and unified way this strange beast that an economy is -partly governed by laws that appear thousands of years old, partly changing and shifting constantly.

$\endgroup$

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.