18 December 2011
The World Bank [has collected](http://rru.worldbank.org/Privatization/) a database of major (i.e. at least one
million USD) privatizations in developing countries from 1988
to 2008. The dataset - which is classified by country, sector and
privatization type - is a fascinating view into the development of
countries such as China, Russia and many states in South America.
I've [imported this dataset into OpenSpending](http://openspending.org/wb-privatizations)
to be able to generate custom aggregates. This is of course risky,
as the list is unlikely to be comprehensive and does not include
smaller privatizations on a local level, which are likely to make up
the bulk of the financial volume.
Some interesting views include the breakdown [by sector](http://openspending.org/wb-privatizations/sector)
and [by nation](http://openspending.org/wb-privatizations/from). But the import into
OpenSpending is problematic in other ways: the dataset does not contain information on
who bought the privatized entities and therefore isn't a generic spending dataset.
I've therefore decided to model it in a reverse mode. The individual entries do
not represent financial transactions but the transferred assets - the source is
given as the source country, while the recipient is the newly formed company.
This may be an interesting preview into the problems OpenSpending will face as
it begins to include balance sheet information.