UN Comtrade logo
United Nations logo


Welcome Guest! To be able to post comments please Login or Register.

Notification

Icon
Error

2 Pages12>
UN Comtrade
#1 Posted : Wednesday, October 17, 2012 1:46:12 PM(UTC)


Level: Admin
Posts: 3

This user forum informs about the upgrade of the United Nations Commodity Trade Statistics database (UN Comtrade) which is currently undertaken by the Trade Statistics Branch of the United Nations Statistics Division of the Department of Economic and Social Affairs (UNSD/DESA). We would like to invite you to provide on this forum any comments or suggestions that you might have concerning the existing or forthcoming database (working title ‘UN Comtrade Plus’).

The main goals for the upgrade of UN Comtrade from the user perspective are the inclusion and dissemination of additional data items according to the new recommendations for international merchandise trade statistics (IMTS 2010) and the inclusion and dissemination of monthly data, to the extent that this additional information is provided by countries (or areas). Moreover, the technology behind UN Comtrade will undergo a very significant upgrade which should provide a greatly improved user experience and resolve some shortcomings that have emerged due to changes in technology since the existing UN Comtrade went online in 2004. Whereas we hope to have a beta version of the new system available in the second half of 2014, we expect that the official version of the upgraded UN Comtrade will become operational in the first half of 2015.

We are looking forward receiving your feedback. Please see enclosed note on how to use this forum. A background note on the upgrade called “Implementation of IMTS 2010 and worldwide data collection of international merchandise trade statistics” which contains information about the context of this upgrade as well as some more details about the upgrade itself has been attached to this announcement. In addition, two technical notes on specific important issues have been posted. The contents of these notes is to be understood as providing the initial framework or plan of this upgrade, which might undergo modifications or amendments as required, i.e. based on the additional feedback from users. Further information on the upgrade and its progress will be provided over time on this user forum.

Ronald Jansen,
Chief, Trade Statistics Branch
UNSD/DESA

 

Edited by user Wednesday, September 18, 2013 5:18:17 PM(UTC)  | Reason: Updated the anticipated release timeline.

File Attachment(s):
IMTS 2010 implementation and data collection.pdf (82kb) downloaded 1,078 time(s).
UN Comtrade interface - first draft.pdf (16kb) downloaded 339 time(s).
Data conversion in UN Comtrade - first draft.pdf (32kb) downloaded 1,124 time(s).
How to use the UN Comtrade Plus forum.pdf (94kb) downloaded 347 time(s).
UN Comtrade
#2 Posted : Tuesday, October 23, 2012 8:19:06 PM(UTC)


Level: Admin
Posts: 3

Some user feedback received ahead of the start of the forum (in particular but not exclusively from members of the Task Force on International merchandise Trade Statistics):

---Received on 26 June 2012---

Some notes on Comtrade…

  1. Does not work with browsers other than internet explorer
  2. In Basic Selection (commodities)
    - Would it not be better to have this as products... commodities can be misleading
    - In the selection, Step 1 (Select source) - Special items seems like a strange name to select level of disaggregation, WITS handles this much better
    - When selecting SITC... you can select 5 digits but not 'leaves'... possible discrepancy between total trade and sum of 5-digit trade
    - My favourite Commodities has very few pre-defined groups (WITS has more)
  3. Reporters and Partners. Would benefit from more pre-defined groupings
  4. Checking data availability. When looking at countries that are below the list one has to keep on scrolling up to see the headings of each code...
  5. When you select reporters you have the individual countries and the group of countries (continents, FTAs, CUs, etc.) in the same list. Would be good to have countries on one side and group of countries on another.
  6. Also, the selection of commodities could be better. It would be great if one can select from a list of products those of interest rather than choosing them by identifying key words.
  7. Sometimes when I submit my query and I am prompted to click on download to download the data the link does not work and I have to restart the whole thing. Even when I save the query and go back to it to download the link might not work. This happened twice recently and once was for a query of less than 50000 lines.
  8. Incompatibility issues with Safari Explorer under Mac are the most annoying. If you want to download you better do it with Internet Explorer under Windows only.
  9. In Reporters/partners precisions should be given regarding “Europe” for people to immediately know if it is Europe 15 or Europe 27? I will eventually know which one to download but not the occasional user. Should be simple enough to separate, they already have Europe EFTA and Europe East as aggregates.
  10. Support for more languages. The website already has tons of information so for people that are not very accustomed with some of the terminology (ex. French students) it can be harder to find what they want.
  11. The central page could be more ergonomic. Having the “general disclaimer” take a third of the page in the middle serves little purpose. Having more space for users to find subcategories more easily could be welcome.

---Received on 9 August 2012---

1. Regarding the conversion (1 : n relationship) between the different HS revisions, an option could be to use the share of the n products in the historical data when the country reported in the converted revision and to apply this percentage during the conversion. It could be helpful that a flag clearly mentions if a product code has been created, removed or reallocated in the HS revision X. This is something that we have implemented in Trade Map.

2. Regarding the memorandum items “XX0000” or “XXXX00” it would also be helpful to have a flag on product codes at the 4 and 6-digit level mentioning that the information is partially missing. This type of metadata is published in Trade Map quarterly or monthly data.

3. ITC fully supports your approach to encourage countries to reduce confidential data at the product and partner level – the present situation creates a distortion in trade data analysis.

---Received on 9 August 2012---

1. First, I propose to identify all relationships (ex. HS2012 to HS2007) 1-to-1 and n-to-1 in all levels (2, 4, and 6-digits). This information will be helpful to ensure that the values of originally reported data (at 2 and 4-digits) must be the same after the conversion, regardless there are non-standard codes (XX000 or XXX00) or not. Of course, even though, this approach will ensure consistencies at higher level (for most of chapters or sub-headings), but it does not solve left-out codes at 6-digits (due to decision to allocate all values to one code during conversion, instead of distribution to related codes).

2. Because the specific groups (UNCTAD’s ICT or Creative Economy as example) are built up from 6-digits, any left-out codes will have immediate impact (break in series). To minimize the impact, currently, ICT groups are built based on reported classification (groups defined in every version of HS), but as scopes are changing in each versions, it does not guarantee perfect continuity. That’s why idea to distribute value in one code to several codes (due to n-to-n or 1-to-n, see example of live horse in HS2012 and HS2007) is quite appealing. I support Christian’s proposal on this. We can discuss further which methods to be applied.

3. In general, I agree to carry over (reported) non-standard codes into converted data. However, for those chapters or headings that change the scope, there is a possibility that those values need also to be distributed to other chapters or headings. Again, with consistency that sum of sub-headings should be less or equal with relevant heading or chapter.

4. From conceptual point of view, putting values of non-standard codes at the next higher level does make sense. However, it is not practical to extract data from different levels and calculate/compare them to find out the existence of non-standard codes. It is easier to just extract all 6-digits on a specific chapter along with its non-standard codes (thus sum of all of those 6-digits [standard and non-standard] codes is equal to chapter). Can we use 98XXXX or 99XXXX for this? By coding XX0000 into 98XX00 and XXXX00 into 98XXXX and disseminating them?

5. The SITC. As UNCTAD relies much on aggregated level of SITC, the loss of information due to non-converted non-standard codes may be significant enough to distort analysis. As point 3 above, can they be carried over?

6. At last, countries need to be aware that excessive confidentiality measure reduces data quality. However, I think that not all countries are aware on how to handle confidentiality properly (retain info in next higher level). It should be mentioned regularly in any workshops / technical assistance.

Markie.UNCTAD
#3 Posted : Thursday, October 25, 2012 9:39:31 PM(UTC)
Level: Member
Posts: 2

I'll start with codification of Taiwan which is currently mapped to code 490 Asia NES.

Background:

Here at UNCTAD, Taiwan Province of China is considered as an economy and is assigned country code 158. Data taken from Comtrade need to be reassigned back from 490 to 158, which in fact might have been diluted already (with a real Asia NES) at partner side. For us it is important, so that country groups must really cover the whole economies (like identifying trade flows to “Developing Eastern Asia” or "Emerging Asian Economies", etc.)

Request: to be able to identify Taiwan separately

Suggested options:

*The best is to keep Taiwan as is (code 158) with the proper UN terminology: http://unterm.un.org/dgaacs/unterm.nsf/8fa942046ff7601c85256983007ca4d8/53ddcc762960b20a85256e98004ab0fa?OpenDocument, if it is not against DESA dissemination policy (whether to be able to disseminate Taiwan explicitly). Here at UNCTAD, we do not have that issue.

*The second is to provide us access to (almost final) data of which code 158 has not yet converted into 490. I’m not thinking tariff line data, as we need data in SITC 3-digit level.

*Third option is to flag dataset where the “real” Asia NES is significant. Therefore, we can better estimate Taiwan based on that information.

Markie.UNCTAD
#4 Posted : Sunday, October 28, 2012 7:23:34 PM(UTC)
Level: Member
Posts: 2

Collection and processing of reported data other than HS

Because many countries disseminate data in SITC or BEC (in addition to HS), I want to suggest possibility of collecting and processing data in SITC by UNSD. With this, data loss due to conversion will be avoided. However, there may be other issues to resolve such as totals might differ.

Thank you

mj
#5 Posted : Thursday, November 01, 2012 2:24:34 PM(UTC)
Level: Member
Posts: 1
why am i not able to access un comtrade's database?
Clarence.L
#6 Posted : Thursday, November 01, 2012 3:32:44 PM(UTC)


Level: UNSD Staff
Posts: 2

Originally Posted by: mj Go to Quoted Post
why am i not able to access un comtrade's database?

The UN Comtrade website was temporarily down.

Currently, the UN Comtrade website and database is up, however, please note that the UN Comtrade website and data may be intermittently down due to temporary disruptions of our servers hosted in New York. We apologize for the inconvenience.

 

Florian.WTO
#7 Posted : Tuesday, November 20, 2012 5:31:32 PM(UTC)
Level: Member
Posts: 2
Location: Geneva
Country: Switzerland

Originally Posted by: Markie.UNCTAD Go to Quoted Post

Here at UNCTAD, Taiwan Province of China is considered as an economy and is assigned country code 158. Data taken from Comtrade need to be reassigned back from 490 to 158, which in fact might have been diluted already (with a real Asia NES) at partner side. For us it is important, so that country groups must really cover the whole economies (like identifying trade flows to “Developing Eastern Asia” or "Emerging Asian Economies", etc.)

Request: to be able to identify Taiwan separately

I agree with this, we also have this issue, especially in terms of partner allocation (e.g. some countries like Saudi-Arabia, UAE and Iran use "Other Asia, nes" quite extensively as "nes-group" with consequences for our data processing where it has to be "re-coded" manually back to "Other Asia, nes" instead of erroneously allocating it to partner Chinese Taipei).

Edited by user Friday, February 08, 2013 9:58:29 AM(UTC)  | Reason: Not specified

Florian.WTO
#8 Posted : Tuesday, November 20, 2012 5:34:27 PM(UTC)
Level: Member
Posts: 2
Location: Geneva
Country: Switzerland

Originally Posted by: Markie.UNCTAD Go to Quoted Post

Collection and processing of reported data other than HS

Because many countries disseminate data in SITC or BEC (in addition to HS), I want to suggest possibility of collecting and processing data in SITC by UNSD. With this, data loss due to conversion will be avoided. However, there may be other issues to resolve such as totals might differ.

Thank you

 

I agree with this. WTO also uses commodity groups based on SITC definitions.

irmavasquez
#9 Posted : Thursday, November 22, 2012 5:26:38 PM(UTC)
Level: Member
Posts: 1

 

What is re-dissemination of Data? Do I need to request a permission for generating graphs or figures with your data in a research working paper? (I refer obviously the source)

Matthias
#10 Posted : Thursday, November 29, 2012 8:58:42 PM(UTC)


Level: Admin
Posts: 2

Originally Posted by: irmavasquez Go to Quoted Post
 What is re-dissemination of Data? Do I need to request a permission for generating graphs or figures with your data in a research working paper? (I refer obviously the source)

Please see http://comtrade.un.org/db/help/PolicyOnUuseAndRedissemination_11Aug2010.pdf for the Policy on use and re-dissemination of UN COMTRADE data. You do not need to request permission for generating graphs or figures in a research working paper.

 

FabienS
#11 Posted : Monday, December 10, 2012 5:24:02 PM(UTC)
Level: Member
Posts: 1
Location: Paris
Country: France, Metropolitan

(i) Something that would be great (for my work at least as I consider network flows), in order to detect the large discrepancies :

In "Others", in the selection of trade flows, the addition of mirror values for Imports and Exports would be great, even if one includes CIF and the other is FOB.

A is the reporter, B the partner.

The mirror value of a commodity imported from B to A is the value of the exported commodity from A to B when B is the reporter, A the partner.

(ii) In the knowledge base: The tools developped by partners is empty. Several links on the working partners (Eurostat, ...) or on their tools or knowledge bases could be added.

A correspondence table on the currencies change rates, by month or year, would be also helpful on Comtrade.

Ele
#12 Posted : Thursday, December 27, 2012 6:24:25 PM(UTC)
Level: Member
Posts: 1
Country: Italy

I need data on export by commodity (SITC Rev 3) for Viet Nam for the year 2011 but they are currently not available. Do anybody can tell me if in the next few weeks is likely an update? Other where else could I find such data?

Thank you very much

jkjung
#13 Posted : Monday, January 14, 2013 12:04:29 AM(UTC)
Level: Member
Posts: 1

Are trade values in USD for the reported year? It might be helpful to have fixed 2000 USD, or even current USD.

MSILtd
#14 Posted : Wednesday, January 16, 2013 1:26:31 PM(UTC)
Level: Member
Posts: 1
Country: United Kingdom

Monthly Data is very useful, thank you, but I would like to more easily download a range of months, rather than typing in 'January 2010', February 2010' etc. In fact, could there be an option for 'All' dates? Or the option to select a range of dates?

Many thanks.

Mikhail
#15 Posted : Tuesday, January 29, 2013 12:37:06 PM(UTC)
Level: Member
Posts: 2
Country: Russian Federation

Error in data.

If you query

http://comtrade.un.org/db/dqBasicQueryResults.aspx?cc=080620&px=HS&r=498&y=1998,1999,2000&p=0&rg=1&so=8

than, very likely, there is an error in row "1999", column "NetWeight (kg)".

Regards, Mikhail

Mat
#16 Posted : Thursday, January 31, 2013 9:05:54 PM(UTC)
Level: Member
Posts: 1

Hi

comtrade is only working on explorer: I confirm I could not download anything with firefox. What a pitty!!! It would be very good if you could make the database open to more browser: the part of explorer is declining steadily, and more and more people do not use it! As Linux user, I cannot use it at all, and have big difficulties accessing comtrade, too bad!!

J
#17 Posted : Sunday, February 10, 2013 8:23:12 AM(UTC)
Level: Member
Posts: 1

The database is not working properly.  When i attempt to select product and countries (basic selecdtion), the database ignores my specific selection and gives me the data on all products within the category and all countries.  This accordingly makes the data huge, and the database is unable to process.  So please, do fix this problem.  The UNcomtrade database is very valuable for our work and for statistical analysis i hope this problem can be fixed soon, though from the forum comments, it looks like it is taking long to fix.

Good luck and your work is very appreciated by us users.

Thanks

Nima Zarrabi
#18 Posted : Friday, February 22, 2013 7:13:25 PM(UTC)
Level: Member
Posts: 1

I have been failing at downloading data since thursday.Whats going on? 

If there is anything wrong, plaese get it sorted ASAP. Im a PhD student who is so in need of your database.

 

Thanks in advance

Nima

Juma
#19 Posted : Friday, February 22, 2013 8:46:04 PM(UTC)
Level: Member
Posts: 1
Location: SPAIN
Country: Oman

I failled to download a data

Umar
#20 Posted : Sunday, February 24, 2013 5:51:24 PM(UTC)
Level: Member
Posts: 2
Country: Pakistan
Is there is any problem in the site as it is not working properly since Friday? if there is any problem, then please try to solve it as soon as possible as my research work is going on. Thanks!
Rss Feed  Atom Feed
2 Pages12>
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.