Data on Sierra Leone

[herbertmcleod]

I am unable to reconcile the data on exports or imports from and to Sierra Leone with what exists in the national accounts or even World Bank data. Could you please point me to the source of your data. Thanks

[alexandersimoes]

The source information for all the data is documented here: http://atlas.media.mit.edu/about/data/sources/

We are using the COMTRADE database for all of the data cleaned by Robert Feenstra's group at UC Davis (for SITC data) and BACII dataset (for HS data).

Was there something specific that you think seems wrong?

[herbertmcleod]

Well for one the export data are very strange. Our manufacturing industry was all destroyed during the conflict and has not yet been rebuilt. So we do NOT manufacture all the goods listed, nor can we think of them being re-experted from somewhere else. I must be missing something, no? Herbert

On Fri, Mar 7, 2014 at 4:05 PM, Alex Simoes notifications@github.comwrote:

The source information for all the data is documented here: http://atlas.media.mit.edu/about/data/sources/

We are using the COMTRADE database for all of the data cleaned by Robert Feenstra's group at UC Davis (for SITC data) and BACII dataset (for HS data).

Was there something specific that you think seems wrong?

— Reply to this email directly or view it on GitHubhttps://github.com/alexandersimoes/oec/issues/13#issuecomment-37037900 .

[alexandersimoes]

The majority of exports appear to be both mining (aluminum, titanium, iron and other ores) and natural resources (cocoa beans, sugar and others), not much in the way of manufactured goods. Are we looking at the same chart? http://atlas.media.mit.edu/explore/tree_map/hs/export/sle/all/show/2011/

If you have access to a more accurate dataset we'd love to include it on the site but for now all of the data we're getting has been provided by COMTRADE. This might be an issue worthy of bringing up with them.

[herbertmcleod]

I was looking at the Atlas and cannot reconcile the information with what I know. We do not produce milk for local consumption but see the export of concentrated milk. That we exported $50 m of glass bottles in one year, when we do not have a factory producing bottles? Am sure that I may not be reading the map well - so please help me understand. Herbert

On Fri, Mar 7, 2014 at 6:22 PM, Alex Simoes notifications@github.comwrote:

The majority of exports appear to be both mining (aluminum, titanium, iron and other ores) and natural resources (cocoa beans, sugar and others), not much in the way of manufactured goods. Are we looking at the same chart? http://atlas.media.mit.edu/explore/tree_map/hs/export/sle/all/show/2011/

If you have access to a more accurate dataset we'd love to include it on the site but for now all of the data we're getting has been provided by COMTRADE. This might be an issue worthy of bringing up with them.

— Reply to this email directly or view it on GitHubhttps://github.com/alexandersimoes/oec/issues/13#issuecomment-37052149 .

[herbertmcleod]

I just tried searching for glass bottles in the UN COMTRADE data base and found 0 items. Can u do the same from your side? Herbert

On Fri, Mar 7, 2014 at 6:22 PM, Alex Simoes notifications@github.comwrote:

The majority of exports appear to be both mining (aluminum, titanium, iron and other ores) and natural resources (cocoa beans, sugar and others), not much in the way of manufactured goods. Are we looking at the same chart? http://atlas.media.mit.edu/explore/tree_map/hs/export/sle/all/show/2011/

If you have access to a more accurate dataset we'd love to include it on the site but for now all of the data we're getting has been provided by COMTRADE. This might be an issue worthy of bringing up with them.

— Reply to this email directly or view it on GitHubhttps://github.com/alexandersimoes/oec/issues/13#issuecomment-37052149 .

[alexandersimoes]

Searching COMTRADE it is true that you find 0 results for Sierra Leone... but the source we are using for the HS data (CEPII) uses various methods for symmetrizing the COMTRADE data e.g. by looking at the countries that say they import from Sierra Leone. This can represent problems as it is not coming "from the horse's mouth" so to speak but often times with data collection you need to do the best you can with what you have. As I mentioned if you know of a more accurate or comprehensive dataset of Sierra Leone's export/imports we would love to include it.

Here's is the working paper from CEPII outlining how they clean their data - http://www.cepii.fr/CEPII/fr/publications/wp/abstract.asp?NoDoc=2726

[herbertmcleod]

This is serious. I was alerted to the work you and Prof Housmann are doing on product space and wanted to examine how we can use it to design the development path for Sierra Leone. We have produce a vision of where the country should be in 2035 complete with socio-economic indicators but without an underlying model to guide us. At the same time we are concerned that the small size of the country could limit growth unless we press on with the sub-regional integration (MRU) that is now beginning to make some progress. The ability to drill down and anticipate what industry is likely to enhance growth outside the current natural resource boom would be very useful. A few people from Harvard felt that Prof Ricardo’s work could be helpful but before talking to him I wanted to understand what the concept of product space is about. Do you think your team and especially Prof Housmann would be interested in a conversation on the matter? Best Herbert M’cleod (Former Adviser in the office of the President and currenty Country Director of the International Growth Centre - Sierra Leone.) On Mar 7, 2014, at 7:00 PM, Alex Simoes notifications@github.com wrote:

Searching COMTRADE it is true that you find 0 results for Sierra Leone... but the source we are using for the HS data (CEPII) uses various methods for symmetrizing the COMTRADE data e.g. by looking at the countries that say they import from Sierra Leone. This can represent problems as it is not coming "from the horse's mouth" so to speak but often times with data collection you need to do the best you can with what you have. As I mentioned if you know of a more accurate or comprehensive dataset of Sierra Leone's export/imports we would love to include it.

Here's is the working paper from CEPII outlining how they clean their data - http://www.cepii.fr/CEPII/fr/publications/wp/abstract.asp?NoDoc=2726

— Reply to this email directly or view it on GitHub.

[cesifoti]

Hi Herbert,

Thanks for your comments and interest in the Observatory. Let me help you a bit with a few clarifications. First, the Observatory at atlas.media.mit.edu is a project from the MIT Media Lab Macro Connections group (my group), not from Harvard. I understand this can be confusing because of a past history of collaboration, but I thought it would be clarifying for you to know who you are interacting with. The Observatory makes accessible data, but we do not produce our data. As Alex mentions, the HS data comes from BACI Cepii, so you should look at that at source, aggregate it at four digits, and see if you have problems with that data as well. Also, please note that trade data is based on what it is reported, not on what it is exported. What I mean by this is that sometimes what it is reported is there because of tax incentives and other forms of kickbacks that people might use to make some quick bucks. Unfortunately, there is no way for us to see if the data that other people collected is the result of misreporting of actual exports (we believe, however, that at the overall level the signal trumps the noise)

The good news is that the Observatory allows you to explore this data inconsistencies. For instance, you are worried about the exports of Glass Bottles. The trade of glass bottles from Sierra Leone all is connected to Nigeria (http://atlas.media.mit.edu/explore/tree_map/hs/export/sle/show/7010/2011/). Is this a reliable export link to consider? I don't know, but the fact that it all goes to one country sounds suspicious to me (usually exports tend to be more diversified in terms of destinations).

Hope this helps clarify some of your questions.

All the best

C

[herbertmcleod]

This is all very intriguing! I will pass it on to the Minister of Finance. We would all like to know who exports $11 m worth of "bottles" to Nigeria. Thank you for clarifying the difference between the Observatory and Housmann’s maps. All the best Herbert On Mar 7, 2014, at 8:58 PM, cesifoti notifications@github.com wrote:

Hi Herbert,

Thanks for your comments and interest in the Observatory. Let me help you a bit with a few clarifications. First, the Observatory at atlas.media.mit.edu is a project from the MIT Media Lab Macro Connections group (my group), not from Harvard. I understand this can be confusing because of a past history of collaboration, but I thought it would be clarifying for you to know who you are interacting with. The Observatory makes accessible data, but we do not produce our data. As Alex mentions, the HS data comes from BACI Cepii, so you should look at that at source, aggregate it at four digits, and see if you have problems with that data as well. Also, please note that trade data is based on what it is reported, not on what it is exported. What I mean by this is that sometimes what it is reported is there because of tax incentives and other forms of kickbacks that people might use to make some quick bucks. Unfortunately, there is no way for us to see if the data that other people coll ected is the result of misreporting of actual exports (we believe, however, that at the overall level the signal trumps the noise)

The good news is that the Observatory allows you to explore this data inconsistencies. For instance, you are worried about the exports of Glass Bottles. The trade of glass bottles from Sierra Leone all is connected to Nigeria (http://atlas.media.mit.edu/explore/tree_map/hs/export/sle/show/7010/2011/). Is this a reliable export link to consider? I don't know, but the fact that it all goes to one country sounds suspicious to me (usually exports tend to be more diversified in terms of destinations).

Hope this helps clarify some of your questions.

All the best

C

— Reply to this email directly or view it on GitHub.

[davelandry]

Closing this issue, as this is not an issue with the OEC itself, but with the raw data we receive.

[herbertmcleod]

Thanks all the same. Herbert

On Wed, Mar 19, 2014 at 10:09 PM, Dave Landry notifications@github.comwrote:

Closed #13 https://github.com/alexandersimoes/oec/issues/13.

— Reply to this email directly or view it on GitHubhttps://github.com/alexandersimoes/oec/issues/13 .