The issue of private sector corruption in Africa
While traditionally corruption in Africa has been seen as a public sector phenomenon, private sector corruption deserves as much attention due to its equally debilitating effects on economic activity.
DURING the nine pre-recession years of the 2000s (2000-08), Africa's income grew by 5.8% per annum and by 3% in per capita terms. Sub-Saharan Africa (SSA), which has been historically seen as a laggard, posted a real GDP growth rate of 4.9% and a per capita income growth of 2.4% during the same period. At this pace, SSA's income per capita was on track to doubling from the current levels of $640 in 29 years.
Even during the global economic crisis, SSA posted a 2% real GDP growth rate, which was higher than in most regions. Africa has staged a quick post-crisis recovery, in major part due to prompt and effective counter-cyclical policy interventions by African governments complemented by sustained external financing (Kasekende, Brixiova and Ndikumana 2010; Ndikumana and Brixiova 2013). The recent strong performance has inspired new optimism and even exuberance vis-a-vis Africa's economic destiny, with some referring to an 'African renaissance' and others even crowning African 'lions' (emulating Asian 'tigers').
At the same time, Africa continues to face critical development challenges. It is host to the highest poverty levels in the world and growing youth unemployment; it faces high vulnerabilities to shocks as its economies are too dependent on a narrow production base. There is a growing consensus that efforts to resolve these structural problems are frustrated not only by lack of resources but also by inefficient use of resources due to poor governance, or more specifically corruption in the public sector as well as in the private sector.
Thus today the problem of corruption has garnered a high level of attention in national and international policy discourse for several reasons. First, corruption causes severe waste and misallocation of financial, human and natural resources, thus retarding growth. Second, corruption suffocates private sector activity and entrepreneurship, perpetuating the dominance of an inefficient public sector, and preventing economic diversification and structural transformation. Third, corruption has substantial distributional effects as it widens income gaps between the rich and politically powerful on the one hand and the ordinary workers on the other hand. Finally, in the globalised world, endemic corruption is contagious through cross-border trade and financial relations. Thus the fight against corruption is a global public good as it leads to a healthier, more transparent and more sustainable international economic order that benefits all economies. These reasons explain the strong impetus in the national, regional and international arena to find effective solutions to the problem of corruption.
Traditionally corruption has been seen as a public sector phenomenon whereby public officials sell political commodities to private actors. While public sector agents play an important role in corruption, it involves two parties in willing-buyer, willing-seller transactions that violate the law in one way or another. Thus, even in corruption deals initiated by public sector actors, private sector actors are not always innocent victims. Private sector operators not only benefit from corruption by securing contracts and business-enhancing privileges, but they also often are the initiators of bribery.
Private sector corruption deserves as much attention as public sector corruption due to its equally debilitating effects on economic activity. Private sector corruption is typically facilitated by weaknesses in the regulatory and institutional framework that make it difficult to monitor the enforcement of rules and fraud deterrent mechanisms. It operates through a number of key mechanisms.
Anti-competitive behaviour in the manufacturing and services sectors
A common feature of the industrial sector in most African countries is the high concentration of ownership often in the hands of family-based networks. These networks are generally connected to the existing and past ruling elite, a key source of access to markets, large government procurement contracts, and preferential treatment in taxation and import/export licensing. The monopolistic privileges enjoyed by large private companies are created and reinforced by regulatory barriers that constrain entry into select sectors. The monopolistic structure of the industrial sector prevents the emergence of new entrepreneurs because incumbency privileges enable the existing conglomerates to undercut prospective entrepreneurs notably using their links to the political regime. As a result, this limits the penetration of new technologies and modern business practices, thereby undermining productivity growth.
In the financial system, corruption and anti-competitive behaviour by private agents can exert heavy costs to society as a whole. Anti-competitive behaviour increases barriers to entry, perpetuating inefficient institutions in the sector. A common form of corruption in the financial sector is insider trading, which allows influential insiders including top-ranking managers in the industry to make profits from moving ahead of the market. At the same time, such speculative trading causes drastic losses for honest investors who are often stuck with undervalued assets.
While access to credit is generally limited in African banking systems, this co-exists with concentration of lending to insiders and connected individuals. In a detailed study on Burundi, Nkurunziza, Ndikumana and Nyamoya (2012) document systematic misallocation of resources resulting from corrupt and nepotistic lending policies whereby managers of banks, their associates and employees have access to credit at below market rates, while outsiders struggle to get loans and pay usury interest rates. Exhibiting a clear bi-modal distribution, the rates charged on loans vary from 0% to 7.5% (with a mode of 4%) for insiders and from 18% to 23.5% for outsiders. Analysis at the account level confirms that in addition to managers and employees, the borrowers enjoying privileged treatment included politically influential individuals. The political leverage of these individuals not only allows them to obtain the loans at cheap rates, but also gives them a licence to default on their loans with impunity. This process of corrupt and nepotistic management of the banking system accounts for the numerous bank failures observed in Burundi (Nkurunziza, Ndikumana and Nyamoya 2012). Thus corruption in the banking sector has important political economy dimensions that explain both its genesis and its persistence.
Corruption therefore can have dramatic consequences on the financial sector as it may lead to financial instability. A recent example is the case of Nigeria where a number of commercial banks nearly collapsed as a result of speculative lending and investment decisions. The culprits in this case were complacent regulation as well as patronage.
In addition to banking fragility, corruption in the banking system creates perverse incentives that promote speculation over real investment. As a result, the banking sector becomes an island of prosperity that is divorced from the rest of the economy. It is not surprising that the banking sector in a country like Burundi would still post such high rates of return on equity (in the range of 25-50% for the dominant banks) when the real sector is suffering from a shortage of investment funds. Bank resources are channelled towards speculative activities, away from the bedrock of the economy - the agriculture and manufacturing sectors. During the 2003-08 period, agriculture and industry received 0.8% and 2% of total bank credit respectively, compared to 67% for commerce. The same occurred in Nigeria where, as Apati (2011) concludes, the banking sector prospered at the expense of the real economy.
In his analysis, Apati (2011) documents how corruption and speculation allowed banks to artificially inflate their worth through creative manipulation of funds and the market. At some point, some of the leading Nigerian banks were estimated to be overvalued by up to 50%. Banks inflated the value of their stocks through speculation, increasing demand for stock, resulting in 'irrational exuberance'. Investigations revealed a number of practices behind the speculative bubble in the banking sector (Apati 2011): (1) use of 'special purpose vehicles' by banks to lend to their own subsidiaries which purchased their stocks; (2) insider trading in banks' own stocks through third-party stockbroking companies; (3) warehousing and dumping other banks' stocks to crush their price; (4) banks setting up shell companies to circumvent the single obligor limit, allowing them to issue large loans to individuals for the purpose of serial purchases of bank stocks; (5) banks using depositors' funds to purchase their own public offerings. All these practices violated banking laws, as a result of inadequate supervision. These problems were certainly not specific to the Nigerian banking system. Any country is subject to similar speculative attacks on the banking sector in the absence of vigilant regulation.
The problems in the Burundian and Nigerian banking sectors demonstrate common features of private sector corruption. First, corruption rides on weak governance and supervision of private sector activity. Managers of banks are able to circumvent the rules to advance their interests, and these practices persist due to ineffective supervision and regulation. Second, corruption thrives on either complacency or complicity of political power. In most cases the bankers enjoy cover from top-ranking politicians who themselves have financial interests in the banks. Third, corruption arises from the actions or influence of former politicians turned bank owners directly or indirectly through bank managers who are their nominees. Apati (2011) tells the story of retired politicians who invest in banks or even create banks that are run by novice nephews and nieces, and how these banks become their avenue to golden retirement.
The costs of corruption in the banking sector are high. Bank failures resulting from corrupt management imply substantial losses for depositors and impose stress on government resources as a result of bailouts. It is therefore clear that rooting out corruption from the banking sector should be on the top of the anti-corruption agenda.
These cases also demonstrate that the focus should not be only on problems of corporate corruption by large multinational corporations operating in Africa. Due attention should also be paid to corruption by domestic firms. The analysis also demonstrates that corruption often entails complicity between private agents and public officials who either provide privileges to the former or turn a blind eye to illicit behaviour by rent-seeking private operators. Under protection by influential politicians, corrupt private actors behave as though they are above the law, benefiting from impunity. This again demonstrates that the political economy dimensions of private sector corruption must be taken into account in strategies to eradicate the problem.
Capital flight, trade mispricing and transfer pricing
Economists have worried about the problem of capital flight since the 1930s if not earlier (Kindleberger 1937). But attention to capital flight from Africa is more recent (Ajayi 1997). Capital flight is understood as illicit movement of capital across borders, motivated by: (1) the need to conceal illegally acquired funds by safekeeping them abroad, taking advantage of bank secrecy; (2) evading taxation on otherwise legally acquired wealth or income; or (3) smuggling of goods and money, including money laundering.
Various measures have been proposed in the literature in the attempt to quantify unrecorded flows of capital using data reported in official documents such as the balance of payments, debt statistics and bilateral trade statistics (see Ndikumana and Boyce 2001 for a review). The most commonly used measure of capital flight is the 'residual method' (Erbe 1985), which is based on the reconciliation between inflows of international financial resources (sources) and the outflows of funds (uses). The key sources are net external borrowing and net foreign direct investment. These resources are used to cover the current account deficit and the remainder consists of net additions to reserves. The residual between sources and uses of funds constitutes the baseline measure of capital flight.
Over time a number of refinements have been brought to the baseline measure of capital flight. First, an additional channel of capital flight investigated is the mispricing of international trade whereby exports are under-invoiced for the sake of keeping the difference abroad while imports are either over-invoiced for the same reason or under-invoiced for the purpose of smuggling. We elaborate on this channel as a source of corrupt financial flows in the next sub-section. Ndikumana and Boyce (2001; 2010) introduce further amendments taking into account exchange rate movements in the calculation of net borrowing and debt forgiveness, as well as adjusting for unrecorded worker remittances.
The existing evidence demonstrates that African countries have experienced and continue to face heavy financial haemorrhage through capital flight. The responsibility for capital flight is equally shared by government officials and private sector operators. From the public sector side, capital flight arises through embezzlement of external debt, including outright theft of borrowed funds that are often directly deposited in the same banks that issued the loans through a 'revolving door'. The public sector is also the scene of other mechanisms that facilitate capital flight including kickbacks and padded procurement contracts on debt-funded projects. Empirical analysis has revealed that a substantial fraction of capital flight from Africa is actually financed by external borrowing. Ndikumana and Boyce (2011a; 2011b) find that up to 60 cents out of each dollar borrowed flees in the same year as capital flight from African countries.
From the private sector, capital flight arises through smuggling of foreign exchange and mispricing of international trade (see next sub-section). The volumes of capital flight from African countries are staggering. It is estimated that from 1970 to 2010, Africa lost as much as $1.3 trillion due to capital flight (Ndikumana and Boyce 2012; Boyce and Ndikumana 2012). Assuming that these funds would have earned interest if invested in the market, cumulative capital flight amounts to $1.7 trillion. This is astounding in a continent that faces large and growing financing deficits and a region that is lagging behind in achievement of national development goals.
The stock of capital flight from Africa far exceeds cumulative official development aid to the continent, and is far more than the debts owed by the continent, making the region a 'net creditor' to the rest of the world. However, while capital flight is in the hands of private Africans and disguised in various forms of liquid and physical assets abroad, the debts are public liabilities and are borne by the African people.
Trade mispricing and money laundering
Another manifestation of corruption in the private sector is the massive volume of money that goes missing through trade misinvoicing, a mechanism which is also used as a means of money laundering to recycle illicit and even criminal money into the formal economy. Trade misinvoicing is a process through which importers and exporters manipulate the reported values of their transactions to reduce the amount of foreign exchange to be surrendered to monetary authorities from export receipts and to inflate the amount of foreign exchange claimed from the authorities to pay for imports. Specifically, exporters understate the value of their export revenues, so as to retain abroad the difference between the true value and the declared value of exports. In turn, importers have incentives for both over-invoicing and under-invoicing: over-invoicing allows importers to obtain extra foreign exchange from the central bank at favourable terms, which can then be transferred abroad. Under-invoicing and outright smuggling allow importers to evade customs duties and regulations.
As a result of trade misinvoicing, the current account deficit will be inflated by export under-invoicing and import over-invoicing; it is reduced in the presence of import under-invoicing. Consequently, the capital flight estimate obtained using balance-of-payments trade data will be too low if the current account deficit is overstated: further capital flight is hidden in the trade accounts. Conversely, the capital flight estimate will be too high if the current account deficit is understated: some of the missing foreign exchange was in fact used to finance unrecorded imports. The net effect of trade misinvoicing can only be ascertained empirically.
The amount of trade misinvoicing is estimated by comparing the African country's export and import data to those of its trading partners as reported in the International Monetary Fund (IMF)'s Direction of Trade Statistics. Assuming that trade data from industrialised countries are relatively accurate, the discrepancy between these and data from their African trading partners is interpreted as evidence of misinvoicing. For each African country, export discrepancies with industrialised countries are computed as the difference between the value of industrialised countries' imports from the African country as reported by industrialised countries, and the African country's exports to industrialised countries as reported by the African country (accounting for the costs of freight and insurance). A positive difference indicates export under-invoicing. Import discrepancies with the industrialised countries are calculated as the difference between the African country's imports from industrialised countries as reported by the African country, and the industrialised countries' exports to the African country as reported by the industrialised trading partners. A positive difference indicates net over-invoicing of imports; a negative sign indicates net under-invoicing.
Trade mispricing implies a double loss for African countries: foregone foreign exchange revenue, and lost taxes on smuggled imports and on under-invoiced exports. These costs are in addition to the direct effect of capital flight and money laundering.
Transfer pricing and tax evasion
The globalisation and increasing complexity in the governance and ownership structure of modern corporations has had dramatic impact on the nature of trade and financial flows. Today a large volume of trade takes place within rather than between large corporations, allowing corporations to engage in 'self-regulated' pricing to improve their bottom line. This complex structure creates a 'corporate veil' facilitating transfer pricing among subsidiaries of the same company that trade with one another.
Along with this increasing complexity of the corporate structure, there has also been an increase in the number and importance of jurisdictions that offer uncompetitive and secretive advantages to corporations, commonly referred to as 'tax havens' or 'safe havens'. While the expression may sound as referring to alien and remote lands, the largest territories serving as safe havens include major cities such as London and New York, with Britain hosting the largest number of tax havens in the world. Indeed transfer pricing takes place right in the centres of legal trade and finance.
The move of economic activity offshore behind the curtain of banking and fiscal secrecy implies heavy costs to African countries. As Nicholas Shaxson points out, it is impossible to tell a complete story of African development without including offshore: 'poverty in Africa cannot be understood without understanding the role of offshore' (Shaxson 2011: 149).
Transfer pricing is facilitated by a number of features in the taxation systems, banking systems and corporate structure, including the following:
1. Banking secrecy, which allows the origin of funds to be concealed, making it difficult for authorities to accurately assess corporate tax liabilities. Banking secrecy also helps in other dealings, including money laundering.
2. Thin capitalisation, which allows a subsidiary of a multinational corporation (MNC) to make high-interest loans to another subsidiary of the same MNC in the same host country. This reduces the amount of taxable profits.
3. Use of 'nominees' to hide the identity of the true owner of assets.
4. Limited or no disclosure of company accounts, which allows inconsistent reporting of corporate transactions and tax liabilities.
5. High-level client confidentiality in the name of banking secrecy, which prevents public scrutiny of trade, financial and fiscal records of multinational corporations.
6. Low or zero tax rates across territories, including tax exemptions, allowing corporations to shift expenses and revenues across jurisdictions, through subsidiaries and shell companies that own intangible assets such as trademarks or provide management services. The preferred destinations of transfer pricing operations are tax havens.
Typically, corporations will use one or a combination of the above features to maximise profits in a legal but illicit manner. They use transfer pricing to shift the costs to high-tax territories and sales proceeds to low- or no-tax jurisdictions. This ultimately inflates costs and reduces taxable profits in high-tax territories while increasing taxable profits in low-tax territories. The trouble is that according to current accounting standards, this creative accounting is ironically 'legal'. But it is nonetheless illicit in the sense that it robs African countries of tax revenue that could have financed vital social needs such as education and health. Even from the perspective of the public in developed countries, these practices are viewed as illicit in the sense that they perpetuate inequality, allowing rich corporations and wealthy individuals to pay no taxes on their incomes while ordinary citizens whose incomes are easily tracked face the full burden of taxation. Hence workers end up subsidising the wealthy's consumption of public goods and services.
African countries host a large and increasing number of global corporations, especially in the natural resource sector, some of which operate on their own while others own shares in national companies. The complex corporate structure, the sophistication of operations in the natural resource sector throughout the entire value chain, and the large volumes involved in the transactions, make the industry fertile terrain for bribery, graft and transfer pricing. This results in substantial revenue losses for African governments, both through sub-optimal rent-sharing covenants as well as through underpayment of corporate taxes.
To illustrate this point, a report by the Extractive Industry Transparency Initiative (EITI) found that while mining companies in Zambia paid $463 million in taxes to the government, there were $66 million of 'unresolved discrepancies' between actual payments and companies' tax liabilities in the same year (Sharife 2011). The same report notes that half of copper exports earmarked for Switzerland never made it there, 'disappearing in thin air'. At the same time, copper price differentials between Zambia and Switzerland were six to one in favour of the latter, implying substantial profits for the companies from copper exports. The flipside is substantial losses for Zambia in terms of foregone tax revenue. It is estimated that the country may have lost up to $11.4 billion in tax revenue in 2008, which is just below its total GDP in that year ($14.3 billion) (Sharife 2011).
Mining tax revenue losses for Zambia are partly due to the fact that companies are able to take advantage of their complex corporate structure and the privileges offered by safe havens to reduce their tax liabilities on Zambian territory. For example, copper mining giant Mopani Copper Mines (MCM), which is a Zambian-registered company, is integrated in a complex web of ownership and domiciliation spread throughout the world. In practice, MCM is almost entirely located in tax havens through its corporate ownership: 73% of the company is owned by Carlisa Investments which is based in the British Virgin Islands, which in turn is 82% owned by Bermuda-based Glencore Finance, which is totally owned by Switzerland-based Glencore International AG. This complex corporate ownership structure provides some insights into the incentives behind the trade circuits of Zambian copper. Given that MCM is a subsidiary of Glencore, it makes business sense that it would prefer to take a low price for its products sold to Glencore even though it could sell for a higher price on the open market. This reduces the profits in Zambia, thus reducing the tax liabilities in the country. At the same time, the arrangement shifts profits to Switzerland where tax rates are lower, thus reducing total tax liabilities at the corporation level.
Transfer pricing provides a convenient route for international companies operating in the natural resource sector in Africa to maximise their profits while evading their key corporate responsibilities. This is in addition to the fact that in many instances, international corporations have negotiated very favourable deals with African governments involving long tax holidays. In the case of Zambia, mining companies generate only 2.2% of the government revenue. However, 95.6% of that is from withheld tax on labour income, with only 4.4% from corporate taxes (Sharife 2011). It is estimated by ActionAid that transfer pricing may be costing the government of Zambia up to œ76 million a year.
Tax evasion through transfer pricing plagues other sectors of African economies beyond the natural resource industry, especially the manufacturing sector whenever large multinational companies are involved. ActionAid reported a vivid example of 'tax dodging' through transfer pricing by the giant corporation SABMiller, which operates in a number of African countries, including Ghana, Mozambique, South Africa, Tanzania and Zambia (ActionAid 2010). In Ghana, SABMiller owns Accra Brewery, which was the first brewery in West Africa, established in 1933.
SABMiller is made up of 465 subsidiary companies in 67 countries, including 65 in tax havens: 17 in the Netherlands, 11 in Mauritius, 8 in the British Virgin Islands, 6 in Switzerland, and 6 in British Commonwealth Dependencies. The group claims to have made substantial contributions to African economies through tax payments. But the company has paid little from corporate tax, taking credit for income tax levied on employees and sales taxes collected by dealers on its products, which are borne by customers. ActionAid estimates that while the group claims to have paid $4.4 billion globally in taxes, only $620 million is from corporate taxes, while 86% is borne by employees and consumers.
Two key techniques are utilised by multinational corporations to dodge taxes: royalty payments to affiliated trademark holders located outside of Africa, and management fees paid also to affiliates and subsidiaries within corporate groups outside the continent. For example, during the 2007-10 period, Accra Brewery paid œ4.6 million in management fees and royalties, amounting to 10 times its operating profits. All these payments were made to two companies belonging to the SABMiller group: Bevman Services AG in Switzerland and SABMiller International BV in the Netherlands. It so happens that both companies are conveniently located in jurisdictions with very low tax rates relative to Ghana. In 2010 alone, Accra Brewery paid œ1.3 million in royalties to SABMiller International in Rotterdam for using its brands (Castle Milk Malt and Stone Lager).
In South Africa where SABMiller also operates, the trademark of Castle, a beer that has been produced for over a century in the country, is registered and owned in Rotterdam. South African Brewery Ltd, one of SABMiller's largest operating companies, pays œ18 million per year in royalties to SABMiller BV in the Netherlands.
Tax evasion results in large foregone government revenues for African countries. ActionAid (2010) estimates that African countries where SABMiller operates have lost as much as œ10 million in taxes as a group due to shifting of royalty payments to tax havens and œ8 million due to payments of management fees to subsidiaries also located in tax havens. In an attempt to put a human face to these tax losses, ActionAid estimated that the lost taxes were 'enough to put a quarter of a million children in school in the countries where SABMiller operates' (ActionAid 2010: 30). For Ghana alone, schooling for about 85,000 children could be supported by the foregone corporate tax not paid by SABMiller's subsidiary Accra Brewery.
Léonce Ndikumana is a Professor of Economics at the University of Massachusetts at Amherst. He is also Director of the African Development Policy Program at the University's Political Economy Research Institute (PERI). He is co-editor of Capital Flight from Africa: Causes, Effects, and Policy Issues and co-author of Africa's Odious Debts: How Foreign Loans and Capital Flight Bled a Continent. The above is extracted from his 2013 paper 'The Private Sector as Culprit and Victim of Corruption in Africa' (PERI Working Paper No. 330). The full paper is available on the PERI website www.peri.umass.edu.
ActionAid (2010). Calling Time: Why SABMiller Should Stop Dodging Taxes in Africa. London: ActionAid.
Ajayi, I. S. (1997). An analysis of external debt and capital flight in the severely indebted low income countries in sub-Saharan Africa. IMF Working Paper WP/97/68.
Apati, Seth (2011). The Nigerian Banking Sector Reforms: Power and Politics. New York: Palgrave Macmillan.
Boyce, J. K. and L. Ndikumana (2012). Capital Flight from Sub-Saharan African Countries: Updates 1970-2010. PERI Research Report. Amherst, MA: Political Economy Research Institute.
Erbe, Suzanne (1985). The Flight of Capital from Developing Countries. Intereconomics, Nov/Dec: 268-75.
Kasekende, Louis, Zuzana Brixiova and L‚once Ndikumana (2010). Africa's counter-cyclical policy responses to the crisis. Journal of Globalization and Development (Symposium Edition), Vol. 1(1), article 16 (January).
Kindleberger, C. (1937). International Short-term Capital Movements. New York: Augustus Kelley.
Ndikumana, L‚once and James K. Boyce (2001). Is Africa a net creditor? New estimates of capital flight from severely indebted sub-Saharan African countries, 1970-1996. Journal of Development Studies, 38(2): 27-56.
Ndikumana, L‚once and James K. Boyce (2010). Measurement of capital flight: Methodology and results for Sub-Saharan African countries. African Development Review, 22(4): 471-81.
Ndikumana, L‚once and James K. Boyce (2011a). Africa's Odious Debts: How Foreign Loans and Capital Flight Bled a Continent. London: Zed Books.
Ndikumana, L‚once and James K. Boyce (2011b). Capital flight from Sub-Saharan African countries: Linkages with external borrowing and policy options. International Review of Applied Economics, 25(2): 149-70.
Ndikumana, L. and J. K. Boyce (2012). Capital Flight from North African Countries. PERI Research Report. Amherst, MA: Political Economy Research Institute.
Ndikumana, L‚once and Zuzana Brixiova (2013). The Global Financial Crisis and Africa: The Effects and Policy Responses. In Gerald Epstein and Martin H. Wolfson (eds.), The Oxford Handbook of the Political Economy of Financial Crises, Oxford: Oxford University Press.
Nkurunziza, Janvier D., L‚once Ndikumana and Prime Nyamoya (2012). The Financial Sector in Burundi. NBER Working Paper No. 18289.
Sharife, Khadija (2011). 'Transparency' hides Zambia's lost billions. AlJazeera.com, 18 June.
Shaxson, Nicholas (2011). Treasure Islands: Tax Havens and the Men Who Stole the World. London: Bodley Head.
*Third World Resurgence No. 309, May 2016, pp 24-29