Econometrics Beat on Twitter

Last week I finally caved in and joined Twitter!

My intention is to use it in tandem with this blog, but we'll see how that works out. I might get totally carried away.

I'll be tweeting @DEAGiles.

© 2012, David E. Giles

More About Spurious Regressions

Students of econometrics are familiar with the "spurious regression" problem that can arise with (non-stationary) time-series data. 

As was pointed out by Granger and Newbold (1974), the “levels” of many economic time-series are integrated (or nearly so), and if these data are used in a regression model then a high value for the coefficient of determination (R²) is likely to arise, even when the series are actually independent of each other. 

They also demonstrated that the associated regression residuals are likely to be positively autocorrelated, resulting a very low value for the Durbin-Watson (DW) statistic. There was a time when we tended to describe a “spurious regression” as one in which R² > DW. 

"Disappearing" Historical Data

It's the bane of my life - and probably that of every other economist who (tries to) work(s) with time-series data. Historical data that simply disappear.

Shazam! Now you see, it; now you don't!

You know the sort of thing I mean, I'm sure. Just when you think you have a nice long, consistent, series of economic data, the statistical agency in question suddenly stops recording it. They change the basis on which the data are gathered (usually for perfectly good reasons), and leave you with a big red DISCONTINUED descriptor.

Gee - thanks a lot!

Sometimes they even "pull" the historical series that you were really counting on, and just provide you with a pretty new series that started yesterday, and won't be of any use to anyone for ages.

Forecasting: Principles and Practice

Forecasting: Principles and Practice is the title of a new book by Rob Hyndman and George Athanasopoulos.

As Rob says on his webpage:

"The book is dif­fer­ent from other fore­cast­ing text­books in sev­eral ways.

• It is free and online, mak­ing it acces­si­ble to a wide audience.
• It is based around the fore­cast pack­age for R.
• It is con­tin­u­ously updated. You don’t have to wait until the next edi­tion for errors to be removed or new meth­ods to be dis­cussed. We will update the book frequently.
• There are dozens of real data exam­ples taken from our own con­sult­ing prac­tice. We have worked with hun­dreds of busi­nesses and orga­ni­za­tions help­ing them with fore­cast­ing issues, and this expe­ri­ence has con­tributed directly to many of the exam­ples given here, as well as guid­ing our gen­eral phi­los­o­phy of forecasting.
• We empha­sise graph­i­cal meth­ods more than most fore­cast­ers. We use graphs to explore the data, analyse the valid­ity of the mod­els fit­ted and present the fore­cast­ing results."

This looks really good!

© 2012, David E. Giles

It's Not Rocket Science!

In fact, it's pretty obvious that this isn't any sort of science.

I'm referring to this little gem, in a post from Eric Crampton in the Offsetting Behaviour blog, back in February.

And it seems that these pseuds. just won't go away - see Eric's post today.

@Students in my ECON 246 course: If there were a Darwin Award for sample surveying, this would win one, hands down!

© 2012, David E. Giles

Happy Birthday, "Your Better Life Index"

One year ago, the OECD released its Your Better Life Index. The Index sought to provide comparative data relating to well-being that go beyond the traditional economic output measures such as GDP.

I analyzed some of the YBLI data in a few posts last year - here, here, and here.

The YBLI has been re-released (updated) today. The OECD tells us:

"Some of the key takeaways from the new version of the Index include:

• No matter which countries people live in, they value the most some combination of  health, education and life satisfaction.
• Men and women who have used the Index value basically the same things.
• The wealthier you are, the more likely you are to make your voice heard in elections, but not by a huge margin.
• Men work more in the labour market and make more money than women, but women are better in other areas, they live longer, are better educated and in most places they are also happier.
• Inequality isn’t just about money, it affects other topics in Your Better Life Index."

BTW, if you have an interest in these issues, then the Freedom and Flourishing blog that Winton Bates writes should be on your reading list.

 

© 2012, David E. Giles

Log Transformations & Forecasting

I enjoyed reading the lead article in the latest issue of Empirical Economics, by Helmut Lütkepohl and Fang Xu. It assesses the quality of forecasts obtained from an ARIMA model that is estimated using the levels of the data in question, as opposed to forecasts that are generated from a model estimated from the logarithms of the data.

Estimating & Simulating an SEM

We all know that structural simultaneous equations models (SEM’s) played a key role in the historical development of Econometrics as a discipline. An understanding of these models and the associated estimators is an important part of our training, whether we use these models or not in our day-to-day work. The issues that they raise have helped shape much of our current econometric tool-kit.

I've posted on this topic before, but here I'm going to look at the results of applying various SEM estimators using the EViews econometrics package. In particular, I'll use a simple well-known structural model to illustrate the estimates that are obtained when different “limited information” and “full information” estimators are used.

Then, I'll take a look at using an estimated SEM for the purposes of simulating the effect of a policy shock.

Complex Survey Data in Econometrics

It's easy to forget that many of the standard results that we learn in statistics are based fairly and squarely on the assumption that the data are obtained by using simple random sampling.

In reality, this is rarely the case. Stratified  and cluster sampling are routinely used by our statistical agencies, and frequently the sampling process is much more complicated than that. So-called "Complex Survey" designs, which involve multi-stage stratification and clustering are very common.

This is something to watch out for in practice.

Access to Public Data

This item, from the Royal Statistical Society's RRSeNews site.

The U.K. Protection of Freedoms Act 2012 now requires public sector organisations to publish open data in a format that is standard and re-usable. Raw data that can't be used without further requests won't cut it any more. Neither will data that are available "electronically", but only in a format that fails to meet open standards.

Goodbye .pdf files, hello .csv files!

This is good news. Of course, there's a kicker - the amendment to the Act allows the public sector to charge for this service.

This development stands in juxtaposition to Statistics Canada initiative of late 2011 (see here), whereby more data are now being offered without charge, but the data aren't necessarily available in a ready-to-use format.

I guess we're making progress here!

© 2012, David E. Giles

And the Winner is......

R will overtake SAS and SPSS in 2015 - according to David Smith in his post on the Revolutions blog.

I can believe that!

© 2012, David E. Giles

A First Lesson in Econometrics

Just in case you haven't previously seen John Siegfried's 1970 piece in the JPE.......... (or here - thanks Dimitriy!)


(H-T to Jia Liu for reminding me of this.)

© 2012, David E. Giles

Re-Writing Economic History

Revisions to official economic statistics are a fact of life. They happen all of the time. Many data are released on a "provisional" basis, and subsequently revised. Sometimes these revisions are quite minor, but there are also times when they can be downright embarrassing.

The 2012 Econometric Game

Last July I posted about the 2011 Econometric Game. The 2012 edition of the Game was held last month, with the top three places going to teams of students from the University of Copenhagen, Aarhus University, and Harvard University.

You'll find a list of the 30 competing teams here.

What a great venture!

© 2012, David E. Giles

EconometricsbySimulation

Francis Smart, a grad. student at Michigan State U., has written to me about his blog - EconometricsbySimulation.

As Francis says, "It is primarily a resource for students who want to learn new methods."

Francis has some nice Stata examples posted, and I'll certainly be following his blog.

© 2012, David E. Giles

Knoema Data Site

Always on the look-out for accessible data, I was interested to see this recent post on the Guardian's DataBlog site. It discusses a new site, Knoema, that was launched last month. They have a blog, that I'm adding to my blogroll.

We'll see how this all develops.

© 2012, David E. Giles

A Good Time to be an Economics Grad. Student?

Last Thursday, Neil Shah had this piece in the Wall Street Journal.

There'a a battle going on between the top U.S. economics departments for new and established talent. Of particular interest to economics grad. students, though:

"Established or not, economists are hot commodities. Last year, the average starting salary for new assistant economics professors was nearly $112,000 – the highest ever in inflation-adjusted terms and one of the highest across academic departments, according to the American Economic Association."

© 2012, David E. Giles

R Videos - and More

Some of us learn easily from the written word, but for most of us some visualization speeds up the process and generally helps with retention as well.

With that in mind I was delighted to see this nice list of free videos that demonstrate the use of R, posted on Ethan Fosse's blog, "Culture, Statistics, and Society".

Also of interest to econometricians will be Ethan's recent post outlining why he's dropped Stata completely in favour of R.

I'll be following his blog, for sure.

© 2012, David E. Giles

Bayes Estimators, Loss Functions, and J. M. Keynes

As a result of my recent post on Bayesian estimation of a simple consumption function, a few people emailed asking for proofs of the results that the Bayes estimator is the mean (a median) [a mode] of the posterior density, when the loss function is quadratic (absolute error) [zero-one].

Let's take a look at this, for the case of a single parameter.

Mathematicians and Economists

Tim Johnson posted a helpful comment on my recent post about Newton. This led tme to Tim's interesting blog called "Magic, Maths and Money".

I particularly enjoyed his two January posts there titled "Why Don't More Mathematicians See the Potential of Economics"; and "Why Don't More Economists See the Potential of Mathematics"

Good questions!

© 2012, David E. Giles

Newton and the Royal Mint

Sir Isaac Newton had a "day job" - Master of the Royal Mint, from 1699 until his death in 1727.

Stressful Times at Statistics Canada

An independent, well-funded, and well-staffed central statistical agency is a prerequisite for rational public policy-making. Statistics Canada is currently under a severe threat.

This is very bad news for those of us who use statistical data to do our jobs, but it's even worse news for the StatCan professionals who are directly affected.

© 2012, David E. Giles

Bias-Corrected MLEs

We all know that the Maximum Likelihood Estimator (MLE) is justified primarily on the basis of its desirable (large sample) asymptotic properties. Specifically, under the usual regularity conditions, the MLE is generally weakly consistent, asymptotically efficient, and its limit distribution is Normal. There are some important exceptions to this, but by and large that's what you get.

When it comes to finite-sample properties, the MLE may be unbiased or biased; efficient or inefficient; depending on the context. It can be a "mix and match" situation, even in the context of one problem. For instance, for the standard linear multiple regression model with Normal errors and non-random regressors, the MLE for the coefficient vector is unbiased, while that for the variance of the error term is biased.

As we often use the MLE with relatively small samples, evaluating (and compensating for) any bias is of some interest.