VoxEU Column Development Institutions and economics Politics and economics

The privatisation of infrastructure: One size does not fit all

Is privatisation of infrastructure a cheap road to development? This column argues that policymakers should recognise that what works for some situations won’t work for all. When it comes to infrastructure, the column suggests that governments and international financial institutions should look beyond the private sector.

he wave of infrastructure privatisations that occurred from the 1980s onwards has led to very different sectoral outcomes across the world (Bortolotti and Siniscalco 2004). For the case of fixed-line telecommunications, in OECD countries privatisations have resulted in higher labour efficiency. By contrast, in non-OECD countries privatisations have mainly increased residential tariffs. Governments and international financial institutions (IFIs) should evaluate under which economic and institutional endowments future privatisations can lead to better performance in infrastructure, particularly in Africa where there are still many public utilities.

The effect of fixed-line telecommunications privatisations on sectoral performance

The outcomes of fixed-line telecommunications privatisations differ widely across the world (see Table 1). Some countries have benefitted from this reform. In OECD countries fixed-line privatisations have led to higher labour efficiency measured by the number of fixed lines per employee. This reform has not affected network expansion though, since these states were already well-supplied. Prior to privatisation, around 50% of the population had access to fixed-lines telecommunications services or, in order words, the bulk of households. In resource-scarce coastal countries in Africa privatisations have also been relatively positive.1 They have resulted in higher labour efficiency as well. However, the increases in residential tariffs to inject capital on the operator have not translated into larger network expansion, even if deployment was very limited before the reform. Indeed, in these countries less than 5% of the population had access to fixed-line telecommunication services.

Other countries have been less impressed by fixed-line telecommunications privatisations. In Latin America and the Caribbean this reform has had no significant impact on fixed-line efficiency, prices, or deployment. However, the worst results have come from resource-rich and resource-scarce landlocked countries in Africa. In these countries, privatisations not only failed to increase outcomes but they actually resulted in a fall in labour efficiency and in some cases in lower deployment. In sum, privatisations in Africa and in Latin America and the Caribbean have failed to expand deployment to reach a larger amount of the population. To illustrate, in Latin America and the Caribbean the number of fixed-lines is only one fifth of those in OECD countries and fixed-line efficiency is about half of that in OECD countries. In resource-rich countries in Africa, the number of fixed lines is as low as 2.2% and fixed-line efficiency is about one fifth of that in OECD. Resource-scarce landlocked countries in Africa have the weakest networks. Fixed-line penetration drops to less than 1% and fixed-line efficiency to half of that in resource-rich countries in Africa.

The impact of fixed-line privatisation on outcomes

 

Fixed-line expansion

Labour efficiency

Residential prices

OECD

Null

Weakly positive

Null

Latin America and the Caribbean

Null

Null

Null

Africa

Strongly negative

Negative

Strongly positive

- Resource-scarce coastal

Null

Strongly positive

Strongly positive

- Resource-rich

Strongly negative

Strongly negative

Weakly positive

- Resource-scarce landlocked

Null

Negative

Null

Note: The outcomes (positive, null, negative) designate the impact of the privatisation of the fixed-line telecommunications operator on the sector performance (fixed-line expansion, labour efficiency and residential prices) Source: Based on Gasmi et al (2011) fixed-effects estimation in 1985-2007 for 23 OECD countries, 25 Latin American and the Caribbean countries and 43 African countries, of which 16 are African resource-scarce coastal countries, 15 are African resource-rich countries and 13 are African resource-scarce landlocked countries.

Should privatisation apply independently of economic and institutional endowments?

There are strong linkages between the performance of privatisation reforms and the countries’ economic and institutional endowments classified according to market profitability, sectoral regulatory power, and country risk. Firstly, profitability in the fixed-line sector is dependent on the existence of an effective demand large enough to cover high sunk fixed costs. Individual willingness to pay in low-income rural economies might not be sufficient (for example, Burkina Faso, Mali). In addition, the emergence of mobile technology with much lower fixed costs and faster deployment rates than fixed-line has acted as an important substitute for voice services.

Secondly, mandated obligations on network expansion have been difficult to impose. Even though regulators both in Latin America and in Africa were created to function separately from the Ministry of Communications, in practice they were not independent from political influence. The management of regulatory agencies was often appointed by political decision makers. In addition, most of the staff of the regulator was composed of personnel that previously belonged to the Ministry of Communications. Even though the wages of the staff in the regulatory agency were set high to avoid interference from private agents, the agencies were not ring-fenced from political interference (for example, Cameroon).

Third and most importantly, some acquisitions were uninformed and risk-averse which led to a freeze in investment under private shareholders (for example, Niger and Rwanda). This last reasoning not only led to stagnation in fixed-line penetration rates, but often also resulted on unvaried residential prices and allocative efficiency. In the case of both Niger and Rwanda, governments are currently investing in new broadband wholesale networks in view of the lack of responsiveness of the private firm to deliver outcomes.

Holding these explanations in mind, we observe that OECD, Latin American, Caribbean, and African countries have widely divergent characteristics on market profitability, sectoral regulatory power and country risk. Between 1985 and 2007 OECD countries were characterised by relatively high market profitability, strong regulatory power, and low country risk. African resource-scarce countries and the Latin American and Caribbean countries instead lagged behind OECD countries particularly concerning regulatory power and country risk. Despite this comparison, both African resource-scarce and Latin American and the Caribbean countries were naturally favoured by their access to international trade flows. African resource-rich countries were strongly penalised with high country risk and low regulatory power and African resource-scarce landlocked countries with very low market profitability and high country risk (AfDB et al 2009). Indeed, resource-rich countries in Africa are particularly vulnerable to episodes related to income redistribution between ethno-regional groups while resource-scarce landlocked countries in Africa are the most prone to anti-growth episodes (Bates et al 2018).

How can governments and IFIs improve infrastructure performance?

Governments and IFIs should consider alternative options to privatisation to increase infrastructure performance depending on the countries’ economic and institutional endowments. This is particularly relevant for resource-scarce landlocked and resource-rich countries in Africa where fixed-line telecommunications privatisation has had a significantly negative impact on the performance of firms. IFIs should consider governments’ public investment options in infrastructure with the arrival of emerging partners such as China and India (AfDB et al, 2011). In the last decade, sources from non-OECD countries such as lines of credit from commercial banks, credit export agencies and, to a lesser extent, foreign direct investment (FDI), have contributed to improve the African ‘savings gap’.

Despite this, many IFIs continue to advocate fixed-line telecommunications privatisation in Africa where there are still many public utilities. The question of under which conditions privatisation of fixed-line telecommunications leads to positive outcomes is therefore timely and relevant. This model might be adequate for those countries that have high degree of openness and reasonable potential demand. In other countries privatisation has led instead to lengthy processes and often to the stagnation of the sector. This holds true both when governments endorsed this policy, for example, Rwanda and Niger and when they did not, for example Cameroon (Recuero Virto 2010).

References

AfDB, OECD and UNECA (2009), African Economic Outlook, OECD Development Centre and African Development Bank, Paris.

AfDB, OECD, UNDP and UNECA (2011), African Economic Outlook, OECD Development Centre and African Development Bank, Paris.

Bortolotti, B and D Siniscalco (2004), The Challenges of Privatization: An International Analysis, Oxford University Press.

Bates, RH, P Collier, BJ Ndulu, SA O'Connell, and CC Soludo (2008), The Political Economy of Economic Growth in Africa 1960-2000, Cambridge University Press.

Gasmi, F, A Maingard, P Noumba Um, and L Recuero Virto (2011), “The impact of privatisation of fixed-line telecommunications operators in OECD, Latin American and African countries: One size does not fit all”, Working Paper, OECD Development Centre, Paris, forthcoming.

Recuero Virto, L (2010), “Open access principles”, 3rd African Union ministerial meeting on Information and Communication Technologies (ICTs) proceedings, Abuja, Nigeria.


1 We use this African sub-division developed under the auspices of the Africa Economic Research Consortium (AERC) to account for the heterogeneous characteristics of the 53 countries in the continent (Bates et al.).

1,155 Reads