Link to new version: pauldavidboll.com/SPUR_Stata_J...

Stata package:
github.com/pdavidboll/S...

The new title reflects that we have significantly extended the practical application section to make this paper more useful to applied researchers generally, regardless of software persuasion.

For sure, but I still think they're both important. E.g., much lower AfD vote share in Niedersachsen and Schleswig-Holstein compared to former DDR, despite being similarly protestant.

Speaking very much as a non-expert, I'd say the first two are the main ones, but all of them to some extent. The main difference to Britain, it seems to me, is that class gradients exist mostly _within_ regions rather than across.

(Class and location/region that is.)

It's very hard for Brits to wrap their heads around those two things not being the same thing.

Related to 3.: Further improvements in e-learning technology, increasing scalability for the "best" educators and displacing the others?

It says "one of the closest", to be precise.

I don't claim that GDP is a perfect measure of human welfare, but this is incorrect. Curing diabetes would free up resources, both of diabetes patients and the health care system, which would be put to other uses (see broken window fallacy). Otherwise we could increase GDP by giving people diabetes?

Would be interesting to know how much of this was the initial "shock" of the spread of protestantism vs the feedback generated from educated mothers educating their daughters and so on.

I don't think "professionals" is the correct translation of "Beamte" in this case. "Beamte" means civil servants, which in Germany comprises many employees of public institutions (teachers, professors, police officers, administrative officials, ...).

#EconSky

Questions and comments welcome !

(n/n)

We present a practical guide to these methods for empirical researchers, along with an easy-to-use Stata package that implements them:
github.com/pdavidboll/S...
(4/n)

Müller and Watson (2024) show that strong spatial dependence ("spatial unit roots") can lead to spurious regression results even with HAC corrections, in a parallel to well-known results from time series. They develop diagnostic tests and a method to remove unit roots. (3/n)

Link to Working Paper:

warwick.ac.uk/fac/soc/econ...

It is well-known that spatial dependence is a problem for inference in regressions that use spatial data. However, standard HAC correction methods are only enough when dependence is not too strong. (2/n)

The hope is for future high speed rail projects to continue in a clockwise fashion back to Old Oak Common, at which point the lines will form a ring, referred to informally in policy circles as the HS25.

The year is 2047, plans for the new HS6 line from Old Oak Common to Walthamstow have been approved. In Walthamstow, passengers will be able to change to the recently opened HS5 line, which will take them speedily on to Woolwich and Bromley.