Talk:Economic statistics

I think this article should just be a redirect to Econometrics, but I'm not sure how to set that up. Anyone able to do it? [00:14, 14 October 2003 Kevinatilusa (Suggested redirect to another page.)] (Preceding [Edit summary] added --TM 18:51, 17 September 2011 (UTC)

Just for the record, they are two very different subjects, and the redirect wasn't really appropriate. -- Avenue 20:58, 17 April 2007 (UTC)

Point of further clarification: From 14 October "Economic statistics" started off only as a redirect to Econometrics. It only became an article in December 2006 per http://en.wikipedia.org/w/index.php?title=Talk:Economic_statistics&action=history. --Thomasmeeks (talk) 18:51, 17 September 2011 (UTC))

Proposal to merge Economic statistics with Economic data in a month

I'd like to propose merging Economic statistics with Economic data in a month to the name of the latter but with "economic statistics" getting equal billing in the Lead. All content would be carried over to the merged article (and of course a #WP:REDIRECT, so anyone searching for "economic statistics" ends up there).

Rationale: They are different names for the same subject. Merger would allow the benefits of synergy from combining them. I believe that the "Econ data" article might also get the benefit of new eyes (of earlier contributors to "Economic statistics") & so improve other parts of the "Econ data".

If there is no consensus against it, the merger would take place in month with a copy of the earlier history of "Economic statistics" & its Talk page left at "Talk:Economic data" for at least month thereafter before archiving.§ Thank you. --Thomasmeeks (talk) 16:28, 28 July 2011 (UTC) § For example, such as what is now at the Talk:Long run and short run archive for the merger of the Short run article into the Long run articles. TM 23:27, 5 September 2011 (UTC)

Oppose Economic statistics is about methods and theory for analysing data, at least as far as the present text goes (e.g. "Economic statistics is a branch of applied statistics") while Economic data is about the data themselves, its sources, publication etc.. Melcombe (talk) 17:10, 28 July 2011 (UTC)

Well, I do respect your perspective, M, which is from the statistics side. I have italicized part of the 2nd sentence at the top to emphasize sharing your concern about preserving that perspective. I approach it from the econ side. I think that the two perspectives are complementary.

Here's how one could rewrite the first (and only paragraph) of the article for use in Economic data (italics added to show the change):

Economic statistics or economic data is a branch of applied statistics and applied economics focusing on the collection, processing, compilation and dissemination data concerning the economy of a region, a country, or a group of countries. Economic statistics is also referred as a subtopic of official statistics for economic data produced by official organizations (e.g. statistical institutes, intergovernmental organizations such as United Nations, European Union or OECD, central banks, ministries, etc.). Economic statistics provide the empirical data needed in economic research, whether descriptive or econometric. They are a key input for decision making as to economic policy.

Only the first sentence changes at all. The 2nd & 3rd sent. above are very close to the usage of 'econ stat' that you attribute to 'econ data'. My impression from reading about earlier economists is that before econometrics (and I) came along, some econ. depts. had courses called "econ. stat." That would have made "Econ. stat." at least as much a branch of econ as of stat. So, there would have been no naming right of Stat. to assert a monopoly claim on "econ stat. I'd make this case too: If econometrics is applied statistics, why wouldn't "economic statistics" as a subject be as much of an Econ subject as in stat.?

The Journal of Literature for library, student, and professional purposes has the JEL classification codes. At JEL classification codes#Mathematical and quantitative methods JEL: C Subcategories, there is "JEL: C8 - Data Collection and Data Estimation Methodology" (with a WP link to Economic data). That sounds very much like what the text of "Economic statistics" was referring to. I have come across some handbooks referring to "economic statistics", which talk about methodology. Does that make it a recognized field within Statistics? There's no evidence of that in the article. It seems improbable that applied statistics as a subject would have tools that apply to only one field (say. econ stat). It's not even mentioned in Statistics. In any case, I don't see why that would preclude its use in Econ data.

I don't know that Economic statistics is in fact described by an authoritative Statistics source as a branch of solely statistics. I could not find it so referred to a such in Britannica. The following article supports your cliam: http://encyclopedia2.thefreedictionary.com/Economic+Statistics. But it's from The Great Soviet Encyclopedia (1979) with the note that "It might be outdated or ideologically biased," so not necessarily a reliable source. If your "oppose" remains such & vetoes the merger, so be it. I just think both articles would benefit from merger. Thank you for your consideration. --Thomasmeeks (talk) 20:14, 28 July 2011 (UTC)

M (or others), I don't know that you would care to comment on this, but I'd appreciate if you did. Below is a further-reworked version of Econ. stat paragraph (w the benefit of one other editor) now in Economic data as the 2nd para.:

Methodological economic and statistical elements of the subject, include measurement, collection, analysis, and publication, whether of individuals or of an economy. 'Economic statistics' is also referred as a subtopic of official statistics produced by official organizations (e.g. statistical institutes, intergovernmental organizations such as United Nations, European Union or OECD, central banks, ministries, etc.). Economic statistics provide the empirical data needed in economic research, whether descriptive or econometric. They are a key input for decision making as to economic policy.

The article obviously still needs a lot more work. But is that para. so bad? Any further thoughts would be welcome, esp. on merger. My thought is that applied statisticians might be able to do a lot with the other material. Thank you. --Thomasmeeks (talk) 19:23, 1 August 2011 (UTC)

• Oppose This is a terrible proposal:

Melcome is correct, and this is not the most difficult topic. Tough for you Meeks if you approach it from the "economics side", because you are wrong, like the JEL codes which mention "data estimation". There are plenty of intelligent economists who don't confuse data with methods of analysis. We can ask for help from the computer scientists and information scientists if need be, to stop an introduction of patent nonsense on WP.

Both articles stink (but are better than candidates for deletion). This discussion is premature until at least one of these is in tolerable shape.

It is not clear that there is anything economic in the description. "Methodological economic and statistical elements of the subject, include measurement, collection, analysis, and publication". All of the topics are statistical issues, which appear also in epidemiology.

Also, the JEL is not a reliable authority because of its "data estimation" nonsense; presumably an intelligent sane economist would have written "data analysis/processing".

The proposal is horrific. This sentence " Economic statistics provide the empirical data needed in economic research, whether descriptive or econometric." is just nonsense. What is non-empirical data? Perhaps a sensible sentence could read "Economic measurements provide the economic statistics that are used [not necessary, since most economic research is junk] in econometric inquiries, in business, public policy, consumer decisions, etc."

This is just a waste of time. It is of course related to Thomas Meeks activities in the last weeks on the Economics Sidebar.  Kiefer.Wolfowitz 19:59, 1 August 2011 (UTC)

What about JEL term "measurement" (based on economic concepts & econometric tools applied to the real-world objects to be measured or probabilistically estimated, not excluding errors in variables)? What about "data analysis" by econometricians et al.

Anyone who looks at merely the footnote above link for "Macroeconomic Data," International Encyclopedia of the Social & Behavioral Sciences, pp. 9111–9117. Abstract & TOC above would correctly infer that from start to finish the article is filled with the methodology of national accounts, that is, national accounting. Similarly for other stats (the unemployment rate, Penn World Table, etc.).

Well, data in made-up examples would be an example in non-empirical data. Statistics-textbook examples are replete with the latter. But please consider following up on your nice ED edit suggestion (or I'll do it otherwise w due credit). Of course, "empirical data" is a common expression, even in scholarly & scientific usage.[1], and, yes, written language is full of redundancies, many of which I agree ought to be avoided. And many should not, given the unavoidable ambiguity of natural language (see above).

I did not know about the Econ stat. article before it was listed in the Template:Economics sidebar. I think that most knowledgeable economists and statisticians with an econ background would agree on merger. Of course, you're incredibly productive in your field of expertise. But you might comprise a censored sample. Merger IMO would lead to symbiotic improvements. It already has in Econ data with the added paragraph from Econ stat. Still, I do appreciate the comment.--Thomasmeeks (talk) 00:17, 2 August 2011 (UTC)

I'm sorry for being short tempered. I am tired of cleaning up a bunch of political articles, and my personal demands are heavy too.

Despite regularly disagreeing with Melcombe, I take his comments with the greatest seriousness. Here you haven't.

A made-up or simulated data-set would not result from measurements, truly. But let us remove "empirical", which is overused in economics, unless artificial data are mentioned. Presumably, made-up data are of interest in economics mainly in forensics (e.g. academic misconduct).

I think that an economic-statistics article could discuss non-official economic statistics and statistical topics avoided in econometrics texts, like survey construction, measurement theory. It should suggest that readers look at econometrics and official statistics for other topics. Perhaps "economic data" could discuss/list sources of data? Have you checked Morgenstern's famous On the accuracy of economic observations, which is well worth reading. (It's a pity that people pigeon hole Morgenstern and von Neumann as conservative, because of their military consulting. They were intellectual radicals!)

Cheers,  Kiefer.Wolfowitz 02:04, 2 August 2011 (UTC)

The comment of 2007 at the top of this talk page looks to want to distinguish "economic statistics" (as an analytical activity) from "econometrics". The article at that time was similar to recent versions of what was here, in not confusing "statistics" with "data". If there is to be a distinction made between "economic statistics" and "econometrics" then, presuming that "economic statistics" is more general (as indicated in the immediately above comment), then something about this should be placed in this article. However the article econometrics makes some sweeping claims for what it does cover and seems to preclude there being anything in "economic statistics" that is not in "econometrics". So these articles need to be reconciled. If necessary the present article could be made into a disambiguation page pointing to both econometrics and economic data. I did a quick Google search, and it seems that most uses of "economic statistics" are actually for the "data" sense ... the exceptions I found were at http://www.amstat.org/publications/jbes.cfm for the ASA's "Journal of Business & Economic Statistics", and at http://encyclopedia2.thefreedictionary.com/Economic+Statistics as a dictionary entry for "Economic Statistics". The latter starts (it is actully quite long) ..."the branch of statistical science that studies the quantitative aspect of economic processes and phenomena in the national economy in conjunction with their qualitative aspect. Unlike more specialized forms of statistics, which study economic processes in particular branches of the national economy, economic statistics studies the national economy as an integrated whole". Melcombe (talk) 09:36, 3 August 2011 (UTC)

Thank you for a sympathetic, nuanced, and practical discussion. A bit more on the above, before 1st noting that perhaps I should have self-censored my "censored sample" comment above. (I do gather, and certainly hope, from the dry sense of humor in evidence on M's page that the user is not likely to have taken offense.) I number below for ease of reference.

1T. That's quite right about the usage of "statistics" within econometrics. The Econometrics article already does relationally distinguish between "econometrics", statistical methods, and economic data at the very beginning. The ED article now first equates then distinguishes [[[semantic]]] usages of "ES" & "ED" in the 1st 2 paragraphs. So, maybe not so much confusion there. Relevant to actual usage in " 'ES' university catalog" Google searches, the "hits" are also, as M suggests, to ES-as-data sources and to course names. On the latter "Economic statistics" is in Econ (not Stat) depts. -- as a prerequisite to Econometrics courses (per 1st 20-30 hits here).

2T. One could certainly argue that theoretical (textbook) "statistics" is broader than theoretical (textbook) "econometrics", but IMO it's semantically somewhat off point per "ES" and "ED" in this discussion.

3T. On the http://encyclopedia2.thefreedictionary.com/Economic+Statistics link above to "Economic Statistics" from The Great Soviet Encyclopedia (1979), I don't think that is very helpful as to English usage in ES. Rather, we might agree that the Soviet Encyclopedia article title (per the quoted definition above) was a Russian variation on 'econometrics'. However, the above ASA "Journal of Business & Economic Statistics" link is a terrific source & might be cited & mined for key words, esp. with WP links for use in ES (then seamlessly adapted to ED). I'd be willing to do that as a "finder's fee" to M in gratitude of the other discussants. Of course, if M or KW were willing to do that instead, that might be better in avoiding annoyance at any lapses in my efforts.

4T.Another semantic piece of evidence is William Kruskal & Judith M. Tanur, ed., (1978), International Encyclopedia of Statistics, v. 2, Index, which lists nothing under ES but does list ED as used in:

a. "Econometrics" article (in "Statistical methods" section, itself evidence per the rest of the article [pp. 188-95], that econometrics is not only statistics), pp. 195-96

b. "Econometric Models, Aggregate" article (in "Forecasting and model testing in practice" section), pp. 184-85

c. "Government Statistics" article (in "Statistical methods" section), pp. 415 ("Gov. stats. classified), 419, 424). (There "GS" is used as a plural, not the singular use of Statistics in the opening sentence: "Government statistics in the modern state are an essential part of a wider information system." In the same paragraph, 'stats' is glossed as "data". Thereafter 'stats' is of course used as a plural in the context of its use as "data".

d. "Index Numbers, Applications" article, pp. 459-60

5T. All this is consistent with M's general finding of ES most commonly and relevantly used in the sense of ED. If ES is best used in a non-ED sense, I'd have thought that there would a corresponding entry in International Encyclopedia of Statistics. There is no confusion there (b/c there's no ES article), and there need be none after merger.

6T. I have no problem with strengthening both articles and have attempted just that, as has now M for ES & KW for ED. Realistically, it is worth noting that before M's edit, the text length of the article was only about a line longer (5 instead of 6) than in December 2006. Effectively now, ES is a shrunken version of ED, since ES has incorporated & adapted all of ES prior to the last edit of ES.

7T. Per (1T) above, again on disambiguation per Econometrics and ES, I'd certainly have no objection to the proposal on its own terms. If such disambiguation was needed to make the merger happen, so much the better.

I regret the length of these remarks and again thank you for your consideration. --Thomasmeeks (talk) 19:29, 5 August 2011 (UTC) [Minor copy edit - TM 00:11, 7 August 2011 (UTC)]

A bit more evidence. The International Encyclopedia of the Social & Behavioral Sciences (2001), like the source at (4T) is where one might expect to find ES listed as an article therein. Instead, there's no such under "Statistics". Rather the only relevant articles under Stat. are:

Data Analysis and Graphics

Econometrics and Time Series

Methods of Data Collection and Quality Control[2] (Per the afore listed "Econometrics and Time Series" article, note that "time series" is in the WP ED article but not in ES) and under "General Methods and Schools" only:

Econometric Software

Econometrics, History of,

Macroeconomic Data.[3] (italics added above for "Data" in the article titles.

In the Journal of Bus. & Econ Stats, the Aims and Scope link to the that journal distinguishes between "statistics" and "econometrics".[4] Arguably there's no need to distinguish if 'Econometrics' is merely a less general branch of 'ES'.

I believe the ES & ED have improved enough in the past month (from suggestions and Edits, certainly of the other discussants, if not mine) to make merger more palatable. Merger is not the same thing as deletion. If there is no merger, everything in ES is equally relevant to ED, & there's reason to keep it out of ED. The fact of no further comments from other discussants may indicate an element of agreement on some points of my previous comment above. I believe that it's very appropriate to oppose a merger on a provisional basis, pending adequate discussion. I'm not at all sure that both or even one of the 2 other discussants are still opposed. Any guidance here would be welcome. Thank you for your consideration. --Thomasmeeks (talk) 20:55, 29 August 2011 (UTC)

Revised proposal as to merger of Economic statistics into Economic data

For ease of reference, I have numbered my paragraphs below. TM 23:27, 5 September 2011 (UTC)

1T. No one except me has commented after my 00:11, 7 August 2011 comment above. I believe I have made enough additional substantive comments in response to earlier comments that no further action would effectively make "no merger" "vote"-driven, not discussion-driven, despite Wikipedia:Polling is not a substitute for discussion.

2T. So, let me here submit an alternative proposal: to make ES a WP:Redirect in 2 weeks in the absence of an expressed majority consensus for no merger if there is no expressed consensus against merger, preceded at that time by such changes as suggested below. This would mot be a full merger, since ES could be reverted back from a WP:Redirect. If there is no consensus for reversion, then a full merger takes place say 6 weeks from today.

3T. I did some intermediate Edits at ED to show what a merge-like ED would look like, first importing the 2 para. from ES, modified to here, as para. (2') & (3'), replacing and supplementing para. (2). So, here's what a merge might look like per the modified para. 2 and new-&-modified para. 3.§ I think that's pretty seamless, but of course anyone might improve it. As of now, everything in ES as of about a month ago is already in ED (modified as appropriate). The latest Edit by M is not reflected nor the 2nd para. of ES, which I added. Thank you for your consideration.

§ My last edit there was to revert back to the Edit before the intermediate Edits. --Thomasmeeks (talk) 21:34, 2 September 2011 (UTC)

You still seem to be hung up on making this article into a redirect, ignoring the better possibility of making this into a disambiguation page: see Wikipedia:Manual of Style/Disambiguation pages. Melcombe (talk) 08:44, 5 September 2011 (UTC)

Note: I have left notes on the Talk pages of the Statistics and Economics projects to try to highlight this discussion to get the attention of others who might want to take an interest. Melcombe (talk) 08:52, 5 September 2011 (UTC)

• I oppose the merge. Procedurally, you can't circumvent policy by fiat, changing the requirement from consensus to lack of consensus for other options. CRGreathouse (t | c) 18:55, 5 September 2011 (UTC)

4T. On the first comment above, any full merger (per below) should have a WP:REDIRECT of ES to ED.# I was proposing Redirect as also a temporary expedient, pending final action on merger. # Yes, disambiguation is an option for no merger, but an embedded link is less obtrusive. Most disambigs that I know of are for significant differences in subject treatment for similarly-named articles. WP is filled with articles that provide different names for what are essentially the same subject. Economic data, which now begins "Economic statistics or economic data", is one such IMO.

# Per the previous section, top Edit, last para. there, a full merge would IMO warrant putting on the merged-to Talk:ED page for at least a month the content of the ES and Talk page article up to the time of Redirect, and the history of the article & Talk page, such as for example at Talk:Long run and short run page at the time of merger and now at same page's archive for the merger of the Short run & Long run articles.

5T. On CRG's Edit above, the procedural point is well taken. Of course I couldn't do a merger by fiat. I have edited my first edit to take account of that. Having remedied the procedural defect above & per (1T), I hope that CRG would consider at least temporarily withdrawing his "Oppose" pending further discussion. So far as the present proposal, no one else has yet voted against it. Raising questions or objections, whether procedural or substantive, is of course quite appropriate. I'd guess that there is no disagreement on the need for responding to substantive points per (1T), not only voting. My big the hope is that the eyes of statisticians would be able the make continued improvements in the ES/ED article. Of course the beneficiaries would be readers who would get the benefits of more than one per perspective.

6T. Per (3T) above, no one has challenged the seamlessness of a proposed-merger link for ED as to para. (2') & (3') there. I believe that the cooperative spirit of the ES article links to the "Journal of Business and Economic Statistics, American Statistical Association and its Aims and Scope as among Econ, Stat, Econometrics, & ES is a good model for this section. --Thomasmeeks (talk) 23:27, 5 September 2011 (UTC)

• I oppose the merge also. There seems to be a complete lack of understanding of the point of disambiguation pages and a complete lack of understanding that a general Wikipedia user might be looking for "economic statistics" in either of the two senses of "econometrics" or "economic data". It is largely irrelevant that some editor has adjusted the "economic data" article to make it look as if these are completely synonymous when to most people not working with "econometrics" and "economic data" they are not. So far there have been three votes against a redirect or "full merge" with "economic data" and one (at the very top of this talk page) against a redirect to "econometrics". Of course we might now get a new section re-proposing a redirect or merge under the pretense that it is proposing something different. Melcombe (talk) 09:12, 6 September 2011 (UTC)

7T. Well, if there is no merger, there is no objection to a disambiguation note at the top of ES noting the alternate usage of ES in the Economic data article. But then a disambiguation page seems unnecessary. The similar ES/ED usage in the opening words of Economic data article and the different Economic statistics usage & link in the 2nd paragraph suggest no need for a disambiguation-page link there either.

8T. It is useful that the previous Edit distinguishes:

"economic statistics" in either of the two senses of "econometrics" or "economic data".

For the record, as noted at the bottom of the (untitled) top section on this page, the present ES entry from its start in 2003 through most of 2006 was not an article at all but a WP:REDIRECT to Econometrics.

9T. Let me, however, quote the ES Edit that changed ES into an article:

21:12, 19 December 2006 Anch02 (Before "economic statistics" was a simply a link redirecting to "econometrics". However, economic statistics is very different from econometrics, see also links.)

A link to the final Edit in that series is here, very similar to the current 1st sentence of the article with one exception:

A branch of applied statistics focusing on the collection, processing, compilation and dissemination of statistics concerning the economy..."

That Edit in its entirety, including links, can reasonably be described as referring to the topic in applied statistics concerned with economicdata. But it would be hard to say why it is "very different from econometrics", if it includes econometrics -- and more. The exception is the recent add, 1st unlinked, then repeated earlier and linked, of:

and (data) analysis[5]

after "dissemination" in the 1st ES sentence. It is those "analysis" adds that arguably turn the article into a statistics-article variant of econometrics+econ stat/data. As an applied subject, 'ES' one hopes would not preclude consideration of ED-relevant material, refs, and links (including external).

Admittedly, the above top-edit proposal of this subsection as stated looks doomed per the 2 week-limit stated at (2T) & comments following. --Thomasmeeks (talk) 23:49, 18 September 2011 (UTC)

Statistics vs. data

• A statistic (singular) is a single measure of some attribute of a sample.
• Statistics is the study of the collection, organization, analysis, interpretation, and presentation of data.
• The terms data, information and knowledge are frequently used for overlapping concepts. The main difference is in the level of abstraction being considered. Data is the lowest level of abstraction, information is the next level, and finally, knowledge is the highest level among all three.
• Econometrics is the application of mathematics, statistical methods, and, more recently, computer science, to economic data and is described as the branch of economics that aims to give empirical content to economic relations.

These are the definitions according to Wikipedia, as I copied them from other articles. Don't know if this helps. Wbm1058 (talk) 15:45, 2 April 2013 (UTC)