A substantial sum of money has been illegally flowing out of Ethiopia during the last decade. What is even more worrying is not just that the levels of out flows are high but also the sizes of illicit capital outflows have been rising at alarming rates. This rather unique pattern has attracted the attention of the general public as well as those of bilateral and multilateral donor agencies.

I will also attempt to put some flesh on the bones of facts presented in the GFI database. I will do so by shedding some light on the political economy context of the illicit capital outflow (IFFs) from Ethiopia.

Stolen money trails

The natural starting point is to get a sense of magnitude on the levels and trends. The GFI data is summarized and plotted in Fig. 1. For the time being we focus on the total flows, that is the heights of each bar denoting sizes of annual illicit money outflows. The sum of the blue and red colors gives total amount of money illegally moved aboard from Ethiopia during that year. This ranged from USD $0.4 billion in 2004 to USD $5.6 billion in 2010.

The average annual outflow was $2.6 billion during 2004 and 2013. This is a sizeable sum of money by any standard. For instance, according to estimates reported by the World Bank, the amount of official development assistant (ODA) Ethiopia received in 2010 was $4 billion but total amount of IFFs during that year was $5.6 billion.

This means in 2010 alone Ethiopia’s IFFs exceeded the ODA it received that year by $1.6 billion. In other words, Ethiopia’s IFFs amounted to diverting the entire aid money of 2010 to foreign banks and then still transfer abroad an additional sum of money.

During the entire period (2004 to 2013) the total amount of money that Ethiopia lost due to IFF was $26 billion. This amounts to stealing nearly $300 per citizen. Alternatively, the size of stolen money was about 11 times the total the amount of emergency aid being sought from donors in the current year to buy cereals from abroad and feed the drought victims.

Potential culprits

One may wonder – who are the culprits responsible for Ethiopia’seconomic fraud at such massive scale? The GFI categorizes possible perpetrators into three groups: (a) financial institutions; (b) complicit business counterparts, mainly importers and exporters; and (c) government officials.

In the Ethiopian case, it is reasonable to exclude financial institutions because there is no foreign bank operating in Ethiopia, and the domestic private banks are extremely tightly controlled. Ethiopia’s most influential banks, the Commercial Bank of Ethiopia (CBE) and the National Bank of Ethiopia (NBE), are owned and run by the government. Therefore, in the context of Ethiopia it is safe to include (a) under (c).

That is to say Ethiopia’s IFF can only be undertaken by importers, exporters or government officials. One would hasten to add that there is a huge extent of overlaps between government officials and big businesses in Ethiopia, since big businesses are highly interconnected with the government and/or they are directly or indirectly owned and run by government officials.

Money diversion channels

Now we can shift our attention back to fig. 1 and consider the breakdowns of the IFFs, the individual component denoted by the blue and red sections in each bar. The GFI applies a methodological framework that accounts for two types of illegal movements of money from one country to another.

The first one is export or import trade misinvoicing. This is measured by using a methodology called Gross Excluding Reversals (GER). This simply mirrors exports by one country with imports of another country and vice versa. For instance, items of imports recorded by Ethiopia should agree with records of exporters to Ethiopia in all aspects – value, quantity and quality.

The second one is various leakages in the balance of payments, measured by using the “hot money narrow” (HMN) approach.The latter one is often referred to as “net errors and omissions” in the balance of payment jargon. For instance, if a donor agency or country recorded $1 million grants to Ethiopia but this does not appear in the records by the authorities in Ethiopia, then the GFI records this as a leakage from Ethiopia’s balance of payment.

It is clear from Fig. 1 that the bulk of illicit money transfer from Ethiopia has taken place using trade misinvoicing, denoted by the blue component of the bar. In 2004, trade misinvoicing constituted only 14% of the total IFFs. In 2013, however, this proportion has grown to 100%, the entire IFFs began to be accounted for more and more by trade misinvoicing. For the entire period under discussion, $19.7 billion (or 76% of the total IFFs) was conducted through trade misinvoicing. The year 2010 is an exception – diversion of “hot money” dominated in that year; it constituted 55% of the total IFFs.

False invoices

Trade misinvoicing can take place in one of the following four ways: over invoicing exports, under invoicing exports, over invoicing imports and under invoicing imports. In Ethiopia’s case, the GFI report indicated import over-invoicing is by far the most important method of transferring money abroad. During the period under analysis, about $19.7 billion was transferred abroad through import over-invoicing.

It is critical to understand how import misinvoicing hurts the Ethiopian economy. This is important in the context of huge public construction projects with substantially large components of imports of machinery and other equipment. For instance, an acquisition of a set of machinery whose real value is $1 million is recorded with inflated invoice of $1.5 million.

The importer allocates project budget at the inflated import value, pays the real value to the supplier and then siphons-off the difference (in this case $0.5 million) and deposits it in a foreign bank account. The real damage to the economy happens in terms of inflated capital expenditure. Perhaps the opportunity large capital projects provide for corrupt officials could be the ulterior motive for the uncontrollable urge to attach such a high priority to large capital projects in economic development strategies.

However, it should be noted that public capital projects are often financed through commercial loans that should be paid back with cumulative interests in years to come. The economic return to capital project would partly depend on the cost consideration at project implementation stage.

The GFI also finds some export trade misinvoicing in Ethiopia’s foreign trade, over-invoicing by $6.5 billion as well as $3 million under-invoicing. In trade based money laundering, the most common types of misinvoicing are import over-invoicing and export under-invoicing. As noted above, the case of import invoicing has no complications – so much over invoicing has taken place and it explains the bulk of trade based money laundering in Ethiopia. However, the case of export over-invoicing is uncommon.

Export over-invoicing do happen although they are rare, e.g. China’s trade with Hong-Kong. Export over-invoicing is required when there is a need to plough back money from abroad and report it as inflated foreign direct investment. This is likely the case with Ethiopia where the authorities have been desperate to report higher foreign investments particularly in the first half of the period under analysis.

Ethiopia’s capital flights dwarfs rest of developing countries

It would prove useful to know how bad Ethiopia’s IFFs is relative to other countries. Fig. 2 below compares Ethiopia with its neighbors, the rest of Sub-Saharan Africa (SSA) as well as the average of developing countries (DCs). The comparison was done by expressing total illicit money outflowas percentage of GDP. The years are grouped into three intervals. For reasons discussed further below, it would prove useful to contrast pre- and post-2005. Accordingly, I have isolated 2004 and then divided the remaining years into two equal intervals.

This revealed astonishing patterns of illicit money outflow from Ethiopia which starkly contrasted with those for other countries. First, throughout the years Ethiopia’s records considerably exceeded those for its two immediate neighbors, Kenya and Tanzania. Second, a comparison of 2004 across the countries shows that Ethiopia’s illicit money outflow was way below the Uganda, SSA, and the DCs averages.

Third, the situation changed dramatically from 2005 onwards. Ethiopia outstripped Uganda, and then closed the gap with the SSA average. Fourth, Ethiopia’s average annual money outflows between 2010 and 2013 reached 11% of the country’s GDP, considerably exceeding the corresponding figures for the other countries – SSA (5%), DCs (4%), Uganda and Tanzania (2%) and Kenya (0.013%). Fifth, it is important to note that illicit money transfers abroad constituted smaller and smaller percentages of GDP for most countries over the years, implying substantial improvements in transparency in their economic management. The situation in Ethiopia sharply contrasts with this reality – illicit money outflow becoming a larger and larger percentage of Ethiopia’s GDP. This indicates transparency in Ethiopia’s economic management has gone from bad to worse over the years.