Thursday, September 20, 2018

Constituency for the 2019 general election - Part 1

Yesterday, the Royal Gazette contained the announcement on the number of constituencies for each province for the long-delayed first election after the coup, to be held early next year (unless delayed again). While this might look like another big step towards getting ready for the election, in real it is nothing but publishing the result of a simple algorithm on how to spread 350 MPs over 76 provinces depending on their population, to make sure that each constituency has a similar number of eligible votes. As the number of constituency-based MPs changed from 375 to 350 as well as the population numbers changed since the last election, there are many provinces which will have a different number of MPs.

But in real this is just the minor part of the preparing the constituencies, running the algorithm is something done in few seconds. The real work is the definition of the constituencies themselves, splitting the area of each province into different parts having a similar population number. Even in those cases where a province has now the same number of constituencies as in 2013, the 2013 definition might have to get modified if the population within the different parts of the province has changed and would make the electoral weight of each vote too much differing. I only hope the commission has already started with that bigger task way directly after it was set up, and not waiting till the last laws for the election were officially signed - as knowing the constituency boundaries is important to start campaigning, or simply for the parties to select the candidates fitting for each constituency.

Wednesday, September 12, 2018

Finding a Muban

I was recently contacted by an adoptee who was trying to find out more about her origin in Thailand. Her translated birth certificate gives the following location
Village 10, Don Cheetuan, Dok Khamtai, Chiang Rai
Obviously, Dok Khamtai district in Phayao province is meant, which already gives the first confusion as that is not in Chiang Rai. That's simply because Phayao province was created in 1977, two years after her birth, so at that time the district was still part of Chiang Rai province.

When looking through the list of subdistrict in Dok Khamtai, there is no Don Cheetuan, the only one which is somewhat similar in name in Don Si Chum. Its name is similar enough to explain Cheetuan as a very badly translation/transcription of Si Chum, and its the only subdistrict starting with Don in all of Phayao province. It also fits as the subdistrict was created in 1972 by splitting off  12 administrative villages from Dok Khamthai subdistrict [Gazette]. Thus the place she should look for is Ban Bun Charoen (บ้านบุญเจริญ), Mu 10 of Don Si Chum - isn't it?

Sadly, it isn't that easy, because the village numbers are not stable identifiers for a location. Not only are administrative villages split when they grow to much populous, what makes it even moire difficult to identify a village is the fact that whenever a new subdistrict was created, the villages in the original subdistrict got renumbered to keep them as a list without holes.

Don Si Chum was split two times, in 1976 San Khong was formed out of three villages from Don Si Chum and another three from Ban Tham [Gazette], and in 1992 eight villages were split off to form Sawang Arom subdistrict [Gazette]. While in many of the Royal Gazette announcements on subdistrict creation the number changes are listed, only since 1993 this is included usually - and for both announcements relevant here it was omitted. Also, usually the village are only referenced by their number, the name is only very rarely added, so the present-day names of the villages are of no help.

When renumbering the villages, there are two possible schemes - either keep the original order and shift all the villages starting with the first hole in the list. Or fill the holes by renumbering only the villages with highest numbers, which has the advantage that less villages get a new number.

If the first renumbering scheme was used, then it would be easy to tell the new number of the village - the original village 10 would now be number 5, Ban Don Lek (บ้านดอนเหล็ก). But I guess the second scheme is the one more likely here. In 1976, the villages 4, 6 and 11 were split off, and assuming that there were still 12 villages in the subdistrict then, the most likely renumbering would have been
  • Mu 9 becomes Mu 4
  • Mu 10 keeps its number
  • Mu 11 becomes Mu 6
 In 1992, the villages 1,2,3,9,12,14,15 and 16 were split off, but without knowing how many Muban there were in the subdistrict at that time its impossible to guess to which number the old 10 was changed. The oldest Muban list with names of this subdistrict I have is from 1998 [Gazette], and that already lists the 10 Muban it has today. Two Muban must have been created between 1972 and 1998, but I have no idea when if they were created before 1992, then Mu 10 might have kept its number again, otherwise the old 10 might now be Mu 1 or 2.

All of the guesswork can only be solved when checking with the full government records, which hopefully the Kamnan or the district office can provide. Or maybe some older people living in the area all their life can remember how the Muban numbers changed. Only other source I could think of are old issues of the Local Directory from the years between 1974 and 1992, which might be found in university libraries in Bangkok. How much easier it would have been if the Muban always kept their numbers...

Friday, September 7, 2018

ccaatt geocode list updated

The Department of Provincial Administration has uploaded a new version of their ID lists, dated from August 30. Comparing them with the previous version from December last year only shows two differences
As no new administrative units were created, its not surprising there were almost no changes. And apparently those new districts in planning did not receive a code yet - unlike in past when many of planned minor districts already had IDs but then weren't created.

Tuesday, August 28, 2018

Na Duang TAO to be renamed

In the meeting number 35 on July 25th, the board to consider draft laws discussed the renaming of TAO Na Duang (องค์การบริหารส่วนตำบลนาด้วง), Na Duang district, Loei province to Kaeo Methi (องค์การบริหารส่วนตำบลแก้วเมธี). Its the first direct change to an administrative entity discussed in these board for a year when the merge of Wang Nuea was approved. Due to bad timing, I cannot read the transcript - is still inaccessible for me from Germany, and during my visit in Thailand recently I downloaded all the PDFs of the past year but this wasn't online then yet. Thus the announcement in the Royal Gazette will show whether the board approved the change or not.

The subdistrict Na Duang (ตำบลนาด้วง) is covered by two local governments, Na Duang TAO and Na Duang subdistrict municipality (เทศบาลตำบลนาด้วง). Thus this name change can either mean that the TAO will get upgraded to a municipality soon, or it simply to clean up the name ambiguities of the former sanitary districts in preparation of the proposed upgrade of all TAO to municipalities.

The name "Kaeo Methi" is the name of village number 5 of Na Duang subdistrict, which is also the location of the TAO office.

Tuesday, May 8, 2018

Elected province governors

The issue why the province governors are all appointed officials sent by the Ministry of Interior and not elected directly by the citizens of the respective province - like it is the case in Bangkok since 1972 - has come back into discussion as it is one of the campaign topics by the newly founded "Future Forward Party" for the forthcoming long-delayed election, as it is mentioned in this interview with the party founder.

The website isaanrecords has already followed up this interview with two articles on this topic. First, the summary of a talk by Tanet Charoenmuang given in Maha Sarakham in April, describes the history of the local governments in Thailand compared to the centrally controlled administration. Tanet was a strong proposer of elected governors for many years, and I really should get back to read more of his book "Thailand - A late decentralizing country" which contains his old publications on such topics.

In "Core arguments for and against elected governors", an anonymous author again states the administrative history of the provinces vs. the local governments by municipalities. Sadly, there is no comment possible at that posting, so I have to place my nitpicking here. The article states that from 1972 till 1994 there one one municipality in Thailand - which is wrong as there were already 119 municipalities in 1972. However, what is true is that Chiang Mai was the only municipality of "Thesaban Nakhon" level after Thonburi and Phra Nakhon were merged, and until Nakhon Si Thammarat was upgraded to this highest municipal level. The other odd statement in the article is that BMA is responsible for four provinces, but in real it is only responsible for the special administrative area of Bangkok, which is something like a province.

Whereas in Tanets talk the Provincial Administrative Organizations are mentioned, both articles don't mention that these local administrations were changed into fully elected bodies after the 1997 constitution. The the issue of elected vs. appointed province governors was in the political debate in the 1990s, yet the powerful Ministry of Interior at that time was able to block these proposals, and to get the topic from the agenda it gave the electorate these rather powerless local governments in parallel to the centrally controlled province administration.

Thursday, April 26, 2018

New Roman Catholic diocese

It is totally quiet with any changes of the civil administrative subdivisions, so to fill the void here its a good distraction that just yesterday the Vatican has erected a new ecclesial subdivision in Thailand. The diocese of Chiang Mai was split, and a new diocese with the seat in Chiang Rai was erected [Vatican news release].

The new diocese of Chiang Rai (Dioecesis Chiangraiensis, สังฆมณฑลเชียงราย) covers the province Chiang Rai, Phayao, Nan and Phrae, and the district Ngao of Lampang province. As only 0.7% of the population are catholic, it is only responsible for 18062 people which in other countries would be just one parish. Joseph Vuthilert Haelom has been appointed as the first bishop, and will take the post officially when he is ordained and consecrated later this year. The Nativity of Our Lady church in mid-town Chiang Rai has become the cathedral of the new diocese.

It was interesting to observe how fast after the news release the English Wikipedia got a first article about it, and quickly thereafter a German and Polish version popped up as well - there seem to be many more active editors working on Catholicism topics than the country subdivisions. The bishop so far only has a German biography, and for the cathedral there's so far only the Wikidata entry I created. I now have to create new maps for the dioceses in Wikipedia as well, those created in 2005 have a very limited resolution, as at that time SVG maps were not yet supported. The only difficulty will be to correctly add the boundary between the dioceses of Bangkok and Chanthaburi, which don't follow the administrative boundaries in Chachoengsao province. But at least of the new diocese I am already done, as you can see by the map embedded in this posting.

Friday, March 16, 2018


The fact that a good graphical display of data is usually more important (and more difficult) than the collection of the bare data was one thing I learned from the talks in the Wikidata Con last year. Sadly drawing things myself is something I have no talent at all, but lately two of the Facebook feeds I read started to post nice infographics almost daily, of which I will present two here.

Entity number as of 2017-12-31
Source: DOPA fanpage
The DOPA fanpage posted the graphical display of the 2017 entity numbers I presented here earlier this month. Of course the numbers are the same, it being displayed this way of course looks more catchy than the plain table I posted. I don't get why they chose to display mountains for the central administrative units, and their height totally not matches the numbers. For the local administrative units, it was of course very difficult to find any icon to symbolize the Provincial Administrative Organizations (PAO) - it's already difficult to explain the function of those to someone not familiar with the Thai administrative system, and guess many Thai don't know about them either.

Labor force survey 2018-02
Source: NSO
The second graphic is from the facebook feed of the National Statistics Office (NSO). Most of the infographics there are of course in Thai only, but at least one they also posted in an English translation - the Labor Force Survey of February 2018. It shows interesting factoids like the numbers of employees by sector (agriculture still dominants) or the unemployment rate by education - interestingly Bachelor degree holders have the highest unemployment. Would of course be nice if they'd post translated graphics more often.

While these graphics are great to catch the eye and give information to the reader, they have one big drawback - they are inaccessible to search engines. Textual data can be easily indexed and then found again later, PDF files are already more difficult especially if the raw text isn't embedded and must the retrieved by OCR, but these pure graphics won't ever be found by Google unless they implement much more artificial intelligence into their crawler.

Friday, March 2, 2018

Entity numbers as of December 31st 2017

To round up the annual statistics announcements, today the Department of Provincial Administration did publish the numbers of administrative units. The numbers are as follows, with the 2016 numbers in brackets if there were changes.
  • Changwat: 76
  • Amphoe: 878
  • Tambon: 7255
  • Muban: 75032
  • PAO: 76
  • Thesaban: 2441
    • Thesaban Nakhon: 30
    • Thesaban Mueang: 178
    • Thesaban Tambon: 2233
  • TAO: 5333 (5334)
  • Special administrative units: 2
The only change at all is that one subdistrict administrative organization is gone, which was the merge of Wang Nua TAO and municipality. And as usual, there is still the odd discrepancy of one subdistrict which is not counted in this statistics, but still included in the population data.

Tuesday, February 27, 2018

Royal Gazette announcement of population numbers 2017

Last Friday, the population numbers for each province were officially announced in the Royal Gazette. The numbers were already online on January 1st on
 as reported here before, so there shouldn't have been any news about this - but oddly there are some differences. Whereas the total numbers are same, the numbers by nationality (Thai or foreigner) differ. These numbers are a bit hidden on, as they are only listed with the age pyramid data (see my blog post). For the whole country, the Gazette announcement states that there are 875,814 foreigners, whereas the age pyramid data states there are only 680,549.

The difference is that there two more rows in that statistics - the people who are currently moving, i.e. have unregistered at their old registrar but not yet registered at a new location. These are 157,722 people. The even bigger number are 722,717 who are registered only at the central registrar (ทะเบียนบ้านกลาง), but not at any local one. Thus most of the "missing" foreigners are only centrally registered, and thus don't show up as foreigners in the age statistics. I now have to think a bit how to best encode this fact in my XML files, and of course also clean up the previous years for which I already imported one of the numbers.

Thursday, January 18, 2018

Population pyramid for Thailand

A German-language news post about a recent cabinet decision trying to increase the birth rate reminded me that I wanted to post about the age distribution in Thailand already. The DOPA statistics page has the age distribution for every year since 1993, and even not just for the whole country but down to subdistrict level. Though my Excel skills aren't that great, I managed to build a automatically calculated graphic.
Age pyramid 2017
One can clearly see in this pyramid that the Thai population is aging, and the number of births has decreased a lot. There are two bulges, one of births around 1970 and a smaller one for births around 1995 - maybe simply those from the 1970 bulge having children at that time. The pyramid is not much different from that of the industrialized countries in Europe.

I am currently adding the national age pyramid data from those DOPA statistics into my XML, and since it is probably easier to use for non-technical users also add them into a spreadsheet. The census reports also include age distribution data, but so far I have only typed in a small part of those - the one from DOPA can be done with copy-and-paste and just a little manual work.

Thursday, January 4, 2018

New historical sites

The new year started the same way as the last year ended considering the Royal Gazette announcements - both the final one and the first announcements by the Fine Arts Department on historical sites.

 On December 28th, the historical site Wat Arun was announced [Gazette]. However, as it was already announced before in 1949, this time it probably was only a clarification of the extend of the protected site, as the original announcement did not include any map, but was just a listing of 36 site names all over Bangkok.

Map of Damrong Phaetyakhon residence
January 3rd a new site was announced, the residence of Phraya Damrong Phaetyakhon (Huat Wirawaithaya) [in Thai: บ้านพลตรีพระยาดำรงแพทยาคุณ (ฮวด วีระไวทยะ)], a building now used by the Operation Center for Displaced Persons (สำนักงานศูนย์ดำเนินการเกี่ยวกับผู้อพยพ) of the Ministry of Interior. As it is not far from the National Library I guess I'll have a photo opportunity next time I'm in Bangkok - I probably have to visit the National Library in order to get any detailed data on the 1919 and 1929 census, so far did not find any other library to have anything.

Wednesday, January 3, 2018

New geocodes

When I yesterday mentioned the geocodes missing in the October update of the ccaatt spreadsheet, I forgot to check whether there is a new version of these spreadsheet online already. Sadly there is now no text-only website anymore which indicates the last update of the tables, so I cannot rely on the web-browser app to automatically check for a new version, but have to download the Excel sheets and check them each time.

It turns out that on December 21st, both the ccaatt and the rcode files were updated, adding the missing geocodes for the new subdistricts of Bangkok and assigning a code to one further municipality.

Tuesday, January 2, 2018

Population data 2017

Directly on New Year, the population data as of December 31st 2017 was made online on the DOPA website, giving the population numbers for each province, district and subdistrict. Sadly, its not possible to link to the actual data for the whole country directly, only for each province - however, the age distribution page also shows the total population numbers.

The total population has increased to 66,188,503, an increase of 256,953 or 0.4%. One percent of the registered population are foreigners. As one can see in my spreadsheet with the numbers at province level, the biggest increase by percent was in Samut Sakhon by 2.1%, in fact all of the provinces around Bangkok gained whereas Bangkok again lost a little.
Population development since 1993
(decrease in 2004 was data correction)

As a little side-note - the population data for Saphan Sung district finally gives the codes for the two new subdistricts which were forgotten in the latest update of the ccaatt spreadsheet - Rat Phatthana got the 104402 and Thap Chang the 104403. I still have to look through the data to check whether any municipality got a new code additionally to those already in the October update.

Another thing I still have to look into is the fact that at least for some months already, the population data no longer stops at the subdistrict level but also includes the administrative villages. However, as this means a tenfold increase of data (and thus download time, and XML file size) this creates some technical problems with my processing - besides the fact that the Muban data is a bit odd in some cases I have looked into so far.