This submission from Publish What You Fund to the IDC enquiry “Investment for development: The UK’s strategy towards Development Finance Institutions” is made in addition to our previous submission on February 3rd 2023.
In light of the recent session with Minister of State for Development and Africa, Andrew Mitchell, we reiterate our call for British International Investment (BII) to publish disaggregated impact/results data for all of its investments.
During the session, Mr Mitchell cited aggregate figures for taxes paid by BII investees and jobs supported as evidence of the impact of BII, and to justify the role of BII in the UK’s aid strategy. This occurs despite BII explicitly stating that these figures are not attributable to the investments that BII makes.
Below, we discuss the case of BII’s investment in Grameenphone to highlight the problems associated with relying on unattributed aggregate impact data to measure the impact of BII’s investments. In short, this approach significantly overstates the contribution of BII, obscures the actual intended uses of BII finance, leads to risks of double counting, and potentially creates perverse incentives for BII investment decisions.
The publication of disaggregate impact data for all BII investments would help to guard against the misuse of such data, and allow stakeholders including FCDO as BII’s sole shareholder, to better understand the impact that these investments have. We therefore reiterate our call for BII to improve the transparency of its impact reporting in line with Publish What You Fund’s DFI Transparency Tool.
BII reports taxes paid by its investee companies as one of its development impact metrics. It does this at the aggregate level and does not report the contribution of individual investments to the totals. BII’s key impact data webpage provides impact data aggregated by sector and country. We analysed the impact data for 2019 and noticed some anomalies. First, the sector contributing the most tax was Communications, which, according to the website, contributed 42% of the total taxes paid. This is in spite of the fact that Communications was not one of BII’s seven priority sectors, and it did not even have its own line in the investment portfolio data presented on their investment data webpage. Second, the country with the highest tax contribution was Bangladesh, with US$1.1bn in taxes paid in 2019, 33% of the total US$3.31bn recorded for the year. This seemed unusually high for a country with total BII investment of US$188m. Kenya, for example, had almost twice as much invested (US$321m) but recorded US$146m in tax mobilised, 7.5 times less than the figure for Bangladesh.
While BII does not provide tax data for individual investments, it had a total of 14 investments in Bangladesh so it was possible to review these and attempt to deduce which project or projects are skewing the figures. Among the companies BII invested in in Bangladesh are several SMEs, the construction of an offshore liquid natural gas plant, an investment in BRAC Bank Limited (financial services for SMEs) and a private health care provider (with the investment focusing on a private hospital). These are all too small to mobilise anything close to the amount of tax claimed by BII for Bangladesh.
Two large-size companies were also listed among the investees – the PRAN Group, one of the largest food and agribusiness companies in South Asia - and Grameenphone, a Bangladeshi mobile phone network. The PRAN investment was a US$30m loan to Emerging World FZC, a company that is domiciled in the United Arab Emirates, which aims to “help boost exports of PRAN Group's food and beverage products from Bangladesh to Asia and Africa”. Since it is not domiciled in Bangladesh and it is an investment in an export-focused company within the group, it is also unlikely to be paying significant sums to the Bangladeshi exchequer.
BII made a US$25m loan to Grameenphone in 2013. This was made directly to the company and Grameenphone is domiciled in Bangladesh. It is, in fact, one of the main mobile phone networks and one of the largest domestic taxpayers in the country. With 54.5 million customers, its annual revenue was US$1.7bn in 2019. US$758m of that was profit, which is subject to Bangladesh’s profit tax. Being a large employer and retailer (selling mobile phones and line rental contracts), Grameenphone also pays large amounts of payroll taxes and sales tax (VAT).
According to Grameenphone’s 2019 annual report, the company paid 85.1bn BDT to the Bangladesh National Exchequer that year. The average BDT/USD exchange rate for 2019 was $1 = 84.7 BDT, so the tax paid was equivalent to approximately US$1bn. This is the only company in BII’s portfolio that could mobilise the amount of tax BII claims is paid by investee companies in Bangladesh. On the basis of this evidence a strong assumption can be made that most of the $1.1bn tax documented by BII for Bangladesh in 2019 consisted of the taxes paid by Grameenphone. This also tallies with the apparently skewed contribution of the communications sector in the sector aggregated tax data.
If this is the case it is problematic for a number of reasons:
BII’s investment in Grameenphone amounts to a $25m loan made in 2013 as part of a consortium of lenders participating in the Grameen Phone V loan package led by the World Bank’s International Finance Corporation (IFC). The total loan amount was $345m, and other members of the syndicate include the German, Dutch and French DFIs: DEG, FMO, and Proparco, as well as the OPIC DFI - OFID. The group also mobilised private financing from Standard Chartered Bank. If BII is claiming all of Grameenphone’s taxes paid, each of the other DFIs could do the same, massively inflating the contribution of DFIs to domestic resource mobilisation.
According to BII the purpose of the loan was to “help modernise the Grameenphone network and increase its coverage in rural areas, acting as a driver for wider economic growth.” This indicates that the funding was for an expansion from the existing customer base into rural areas and for upgrades to the existing network. Since the loan was for these specific investments it is likely that the core business of the company could have continued to run without the additional finance. Given this, the claim that the $1bn paid in taxes by Grameenphone is a development impact from BII’s investment is highly dubious.
Measuring development impact in this way would create all sorts of perverse incentives if decisions are subsequently influenced by these results. It would discourage investment in companies that are not paying large amounts of tax, including those that aren’t highly profitable such as energy or infrastructure investments, and loss-making companies in the early investment phase that may need finance in order to become profitable. On the other hand, this approach could create incentives to invest in the highest domestic taxpaying companies in a country. These are often not the companies most in need of development funding and they could likely raise finance from the private sector.
The above case highlights the problems associated with using unattributed aggregate impact data to convey the effectiveness of BII’s operations. The fact that in 2019 approximately one third of the total tax paid by BII investees was made by a single investee is obscured in aggregate data. While we believe we have correctly identified the investment involved (and this has been confirmed by BII staff) stakeholders should not have to go through a process of deduction to reach these conclusions.
Moreover, the lack of disaggregated data obscures information relevant to stakeholders. During the session with Mr Mitchell, questions were raised about why the taxes paid by BII investee companies had fallen significantly in recent years (these fell from $3.31bn in 2019 to $1.51bn in 2021, a 54% decline). In our view, these questions were answered unsatisfactorily. The main reason taxes paid have fallen is because BII’s investment in Grameenphone has now closed, so taxes paid by the company are no longer included in the aggregate data. Disaggregate data would make this immediately apparent to stakeholders.
Finally, BII makes widespread use of financial intermediaries in the form of private equity funds that are often domiciled in offshore financial centres. While this may be an appropriate use of resources, it inevitably raises questions about where and how much tax is paid by direct and indirect investees of BII. Aggregate data obscures this important question, which was again raised by the IDC panel.
As evidenced by our research into Grameenphone, commercial confidentiality concerns should not apply to the publication of disaggregated tax data as this information is often publicly available through other sources. Similarly, BII should make the ability to disclose key performance indicators such as taxes paid and jobs supported, a condition of investment.