A Independence and Central Planning and Management, 1975–1986

In 1975 the official public accounting system was based on the colonial frameworkFootnote ³, namely the Public Accounts General Regulations of 1881, the Finance Regulation (Regulamento da Fazenda) of 1901, and later legislation updates (Portarias dated 1947 and 1958). However, the exodus of Portuguese settlers and the very low level of education and experience with administration among Mozambicans at independence entailed significant challenges in managing the system. The point of departure for Mozambique in relation to PFM was a rather antiquated and little transparent system that had been managed by Portuguese bureaucrats, many of whom left the country at independence, not bothering to hand over or train their successors, sometimes even sabotaging the systems on their way out. Government accounts from the late 1970s were only reconciled and audited in the early 1990s with Swedish technical assistance.

During the first decade of independence, until the end of 1986, the PFM system was adjusted to serve Frelimo’s objectives of dismantling the colonial administration and promoting fast socio-economic development in the context of a Marxist-Leninist ideology (GoM 1981). At its 3rd Congress in 1977, Frelimo decided to implement an economic system of central planning and management (Frelimo 1977a, 1977b). The state planning system controlled the public financial system, the monetary system, wages and prices, and the material balance system. These institutional changes were driven by popular ideology distancing the new and sovereign Mozambique from its colonial past under a leadership that enjoyed widespread popular support. It involved the Mozambican population in a limited manner, but much more than any leadership had done before. Issues like transparency of public accounts were not at the centre of debate; it was about fuelling social transformation and increasingly how to finance the defence against an insurgency supported by the hugely superior neighbouring countries, Rhodesia and South Africa.

To improve internal revenue sources, the tax code was reformedFootnote ⁴ in 1978, increasing the weight of direct taxes, in line with a progressive system (Byiers Reference Byiers2005; Theodossiadis Reference Theodossiadis2004: Table 8.1). Especially taxes on company profits increased. In the category of indirect taxes, which was the major component of revenue (Table 8.1), unessential goods were subject to higher taxes than essential goods. This reform was successful for a few years, when the share of total current revenues over the GDP increased from 11.1 per cent in 1978 to 14.2 per cent in 1983 (CNP 1985; Sulemane Reference Sulemane, Rolim, Franco, Bolnick and Andersson2002).

Budget allocations served the purpose of promoting fast industrialisation, the collectivisation of agriculture, the expansion of state enterprises, and the expansion of primary education and health services. As the massive costs of the war in the 1980s aggravated,Footnote ⁵ a higher proportion of expenditures were allocated to military purposes (‘other’ expenditures in Table 8.2). The informal sector became gradually larger in the 1980s, reducing the share of revenue to GDP.

The share of total current revenues over the GDP declined from 14.2 per cent in 1983 to 8 per cent in 1986, and the overall deficit after grants as a percentage of GDP increased from 10.1 per cent in 1982 to 15.3 per cent in 1986 (CNP 1985: 40; World Bank 1989; Sulemane Reference Sulemane, Rolim, Franco, Bolnick and Andersson2002: 50;). The public debt including arrears rose to US$3.2 billion in 1986 (World Bank 1987).

Overall, the centralised economic model, including its PFM component proved inefficient. Viable companies were subject to foreign exchange restrictions and many public companies were loss makers. The violent conflict was a huge burden on the economy and Mozambique became heavily indebted. After a few years, the economic system collapsed and forced the government to switch to a market-oriented economy in 1987 and to move towards a democratic political system in line with conditionality of the international creditors and donor community.

B Economic Rehabilitation and the Market Economy, 1987–2009

We highlight three sub-periods in PFM reforms between the late 1980s and the discovery of natural gas reserves in 2010. The first started in 1987, when the government implemented a wide-ranging series of reforms aiming to shift the economy towards greater reliance on market mechanisms. Although the initial economic and PFM reforms were successful, in terms of reducing the public deficit as a share of GDP from 10.1 per cent in 1986 to 4.7 per cent in 1989 and recovering national production (8.1 per cent GDP annual average growth rate in 1987–89), the country remained at war under the one-party political regime opposed by Renamo. It was only after the multiparty and democratic elections in 1994 onwards – marking the beginning of the second sub-period – that the PFM reforms could proceed more systematically, in a more coherent and integrated manner.

The volume of internal revenue and foreign aid rose substantially, allowing for the creation of the State Financial Administration System (SISTAFE) in 2002, signalling the beginning of the third sub-period. Significant advances took place in the form of tax reforms, the creation of the Tax Authority (AT) in 2006 and the development of various budget cycle subsystems. However, the positive dynamics of PFM reforms, supported by foreign aid in the context of sustained economic growth and poverty reduction, changed when the existence of large reserves of natural gas was confirmed in the Rovuma Basin in 2010. Since then, government priorities have been focused on minerals and other natural resources. Before we discuss this shift, we describe the three stages in more detail.

C Extraction of Natural Resources, 2010–Present

The period since 2010 saw a shift in the PFM reform effort with the confirmation of the existence of natural gas reserves in the Rovuma BasinFootnote ¹⁰. The prospects of profiting from the extraction of hydrocarbon resources and other mineral resources prompted Frelimo and its governments to shift focus from PFM reforms and prioritise instead the quest for opportunities in mega projects related to mineral resources, especially oil and gas.

The inaugural message from donors to President Guebuza on the start of his second mandate was a freeze of general budget support, triggered by irregularities in the 2009 parliament and presidential elections. The decision came as a complete surprise and enormous disappointment to the government. The G19 had rarely been able to make radical decisions and the government found comfort in the fact that uniting 19 donors, some with very different agendas in Mozambique, on anything but status quo would be very difficult. The government also had informal observers in the G19 secretariat and some of the technical working groups who would catch any discords of significance. This time, however, a good handful of ambassadors knew that the geopolitical winds had changed and headquarters would welcome a harder line, including possible savings that could be spent in fragile and post-conflict states. They acted quietly and without support staff during the big Mozambican holidays after the New Year in 2010, and skilfully organised the motion in a way that would leave softer donor voices behind in the policy dialogue with the government as well as with a problem explaining headquarters why they were lenient and not positioned where the action was.

The event exposed Mozambique’s vulnerability to external shocks, and although ideas were certainly maturing already, the experience strengthened the Guebuza administration’s resolve to reduce donor dependency and move future revenues from gas forward. Moreover, Frelimo tightened the control of the state apparatus, and allegiance to the party became central, sometimes at the cost of a law-abiding behaviour. A series of laws and regulationsFootnote ¹¹ were put in place from the 1980s through the 2000s to dissuade civil servants at all levels from abusing public goods or benefiting illicitly from using public goods. Furthermore, it was expected that the investment in ‘sophisticated, modern and advanced’ electronic and informatics systems, with international and advanced PFM models with the audit and report mechanisms would increase the efficiency of the system and make it easier to control the misuse/abuse or illicit use of public funds. A further reference can be made to the Probity law (PoM 2012, Law 16/2012), which explicitly forbids the abuse of party powers in relation to the functioning of the state entities. However, reality was that these laws were regularly ignored and the practice of asking for bribes in relation to public tenders started escalating.

The first consequence of the above change of priorities was the decision by the government to issue state guarantees for the so-called Proindicus and MAM loans, not included in the budget submitted to the parliament for approval, and which only became publicly known in April 2016 (Hanlon Reference Hanlon2016a, Reference Hanlon2016b; IMF 2016b; Wirz and Wernau Reference Wirz and Wernau2016). They are widely referred to as the ‘hidden debť scandal and, together with the EmbraerFootnote ¹² case, reflect unlawful behaviour by high government officials (IMF 2018; Melo Reference Melo2018; RM 2021).

A second consequence of the change of priorities was a neglect of the factors that could keep peace with Renamo and prevent the economic speculation on the coast of Cabo Delgado that fuelled the eruption of armed conflict in the North and the occurrence of various waves of high-profile kidnappings (Hanlon Reference Hanlon2018, Reference Hanlon2021a, Reference Hanlon2021b; Orre and Rønning Reference Orre and Rønning2017; Pitcher Reference Pitcher2020).

A third consequence is a generalised lower quality of public services, and increasingly evident corruption. To illustrate, Mozambique fell to 149th position out of a total number of 180 countries in the 2020 corruption perceptions index (Transparency International 2020). VAT refund delays and the excessive bureaucracy and non-transparent procedures that taxpayers face when dealing with the tax authority created further opportunities for corruption and abuse of power (Cortez Reference Cortez2014; IMF 2019b).

Put differently, while the PFM reform programme continued in the 2010s and Mozambique was declared compliant with the EITI in 2012 (EITI 2020), this happened in a political environment that increasingly jeopardised both economic and institutional progress. As a result of the 2014 general and presidential elections, Filipe Nyusi replaced Armando Guebuza as the new President in 2015, becoming later in the year also President of Frelimo. The Ministry of Finance and the Ministry of Planning and Development were merged in the Ministry of Economy and Finance in 2015. Planning and Budget functions were merged again into the same directorate. As the public became aware of the ‘hidden debť decisions taken in 2013–14, the group of nineteen donors interrupted all budget support in 2016, and the IMF interrupted its policy assistance. Mozambican public and international partners lost confidence in the country and in the PFM reforms.

Contradictions between Frelimo and Renamo intensified, and the Constitution of the Republic was amended in 2018. The declared aim was to advance the decentralisation process (PoM 2018). However, introducing the position of provincial state secretary has also been interpreted as a move to take away key political and management functions from the elected provincial governor on the one hand, and as Frelimo’s pre-emptive action should opposition parties win the provincial governors’ positions on the other. Should Frelimo lose in local elections, the party would keep political control in provinces through the provincial state secretary.

While presidential, parliamentary, and provincial elections took place in 2019, they were controversial. In parallel, specific PFM reforms continued. The SISTAFE law was revised (PoM 2020) and a pilot planning and budget module and social and economic planning and state budget simulation were carried out in 2020.

A Building a PFM System: Successes and Constraints

Having outlined the main events that influenced the PFM system and marked its different periods, we now put this analysis in perspective and evaluate the PFM reform process. During the period 1975–86, the PFM system was adjusted to the purpose of national revolution, Marxist-Leninist ideology, and war, within a political framework of a one-party state. The government successfully managed the transition from the colonial period to post independence years and started implementing a development strategy based on the industrialisation model of the socialist countries led by USSR. Until 1980–81, the economy and state revenues were under control. There was popular support for the political and economic system adopted by Frelimo government, and political leaders were trusted. They were the national liberators and were adopting policies favouring rural and lower income families. Also, there was no evidence of private enrichment using public resources.

After 1981, Mozambican GDP and state revenue started declining each year, until 1986. The economic system based on central planning and management, the failure of the socialist countries to replace western countries and the rest of the world as commercial and investment partners, and the ongoing war were the main factors for the economic decline, and the PFM system started failing. The lack of separation of powers and non-transparent management prevented external and internal auditing, efficient budget management, effective revenue collection, and appropriate fiscal policy analysis. The Popular Assembly, Supreme Popular Court, CNP, the Ministry of Finance, sectoral ministries, and other public entities were unable to critically identify and correct the public financial system and management distortions for about a decade.

During 1987–94, the PFM system changed substantially, adapting to the market economy and becoming more efficient and transparent. Reforms took place through the approval of new legislation, the change in administrative structures and procedures, and training. Accordingly, this period saw key general reforms that contributed to overcome many PFM system constraints, a new 1990 Constitution, the peace agreement in 1992 and the first democratic and multiparty elections in 1994. However, reform efforts were initially constrained by the war, the lack of separation of powers, a weak rule of law, and the political system with a single party. Some of this was part of the underlying deal between Mozambique and western donor countries.

The sequencing of the reforms introduced by the Bretton Woods was less than optimal in economic terms, as the liberalisations of prices happened at a time where no supply response was possible due to the war. However, the early onset of economic reforms, before peace and armistice, allowed Frelimo to gain and maintain control of the reforms, including privatisations, before Renamo got demobilised and installed as opposition party. The protracted peace process from 1992 to 1994 also reduced the likelihood that Renamo could reject defeat at the first multiparty elections, as farmers, including Renamo supporters, and indeed Renamo’s own rank and file combatants had settled and gotten used to a much more comfortable life than bush fighting and queuing for food in refugee camps.

The sequencing of reforms also meant that the government maintained control of the banking system during this first phase of fiscal reforms. Privatising the banking system was difficult and slow because the state banks were not commercially interesting to foreign banks. The same could be said for the Mozambican market until a credible peace was achieved and economic growth had been sustained for some time. As the PFM-reforms progressed and subsidies were exposed and subsequently suppressed, much of the problem with subsidies to loss making state entities moved to the banking system through loans to companies that would never become viable entities in a market economy. The subsequent cost to the economy of this mining of the banking system has been felt for decades in Mozambique. Moreover, as the fiscal subsidies moved to the banking system as unrecoverable loans, it increased net credit to the economy just as much as it would have if handed out from the budget. This limited the expansion that otherwise could have been accommodated on the fiscal side, including in the social sectors, where expenditure, such as salaries to teachers and nurses in rural areas, was much more likely to generate strong multipliers than the loans to state corporations under privatisation.

B A Feverish Frenzy of Initiatives

In 1994–2009, the PFM system started to change and systematically reformFootnote ¹³ in what can best be characterised as a feverish frenzy of initiatives. There was political commitment at the highest levels within Frelimo and Government, expressed through the main government programmes and budgets. The Ministry of Planning and Finance (1995–2004), and from 2005 onwards, the Ministry of Finance and the Ministry of Planning and Development were in charge of reforming the PFM system. The government was in agreement with international funding partners, bilateral and multilateral, such that aid financing was allocated to PFM reform programmes, technical assistance was provided, and formal and on-the-job training was carried out.

The tax administration body was reformed over time concluding in the establishment of the Tax Authority in 2006. Tax revenue collection capacity increased substantially (Figure 8.1). The budget reforms completed the cycle components, by improving the formulation, approval, execution, accounting, and external audit. Civil society gained access to budget information and had the opportunity to publish independent budget analyses. Internal audit and procurement improved.

Figure 8.1 Tax revenues net of ‘capital gains’ increased from 2006 to 2011 (share of GDP)

Source: Authors’ illustration based on IMF (2007, 2009, 2010, 2013, 2016a, 2019a).

The TA contributed to improving the recording of external funds entering the country for financing investment projects.Footnote ¹⁴ As a weakness, TA was auditing the State General Account (CGE) up to two years after the close of the fiscal year at the beginning, 1998–2002 (Hodges and Tibana Reference Hodges and Tibana2005: 65). With the SISTAFE law this deadline was shortened, making parliamentary scrutiny of the budget execution more meaningful.

The multiparty composition of the parliament led the executive to respond to the public representatives about the management of public resources. However, the parliament remained quite weak in effectively overseeing the executive. Frelimo maintained absolute majority in parliament and most opposition politicians did not have experience with or technical insight to really scrutinise fiscal management and challenge the government. The limited parliamentary oversight was also in part due to the high proportion of external aid in the budget, combined with the fact that aid funded projects and programmes were discussed directly between donors and the executive. Moreover, big parts of the aid funds were regularly off-budget. Other limitations included the fact that tax rates were not part of the budget legislation submitted to the parliament, and by law, the executive had a wide latitude to reallocate the expenditure after parliamenťs approval of the budget (Hodges and Tibana Reference Hodges and Tibana2005: 9–10).

Civil society gained more space for monitoring the executive within the multiparty democracy environment. Nevertheless, in the late 1990s and early 2000s civil society remained weak in demanding information on the management of public resources and in checking the legality of executive actions. One cause for this weakness was the lack of interest due to the narrow domestic tax base of the economy. Another factor was the low analytical capacity to analyse budget issues.

International partners played a significant role in the PFM reforms, investing resources, providing technical assistance and advice, as well as close monitoring. PARPAs, many public programmes and projects, and the respective reports were elaborated with close donor involvement. In 2004, fifteen aid partners signed a Memorandum of Understanding with the government to monitor government performance, decide their disbursements, and plan the application of funds in two annual joint sessions, the Midterm Review and the Joint Review (Batley et al. Reference Batley, Bjornestad and Cumbi2006).

The World Bank also collaborated to a certain extent with the government in preparing their credit programmes and projects and carrying out assessments and studies, like the Public Expenditure Review and the Country Policy and Institutional Assessment (CPIA). The IMF used to carry out at least three missions per year to Mozambique under the various programmes, like the Policy Support Instrument or the previous Poverty Reduction and Growth Facility, and also under the Article IV when an IMF team discusses with officials the country’s economic developments and policies. In these missions, the IMF considered as structural benchmarks some of the PFM reforms, including the e-SISTAFE roll-out for all budgetary operations for goods and services to twenty-two additional ministries and organs in 2007 (IMF 2007). In many ways, these reforms were very successful, which can be seen not only from external technical evaluations, but also from the dynamics of political economy and relations with donors. When rent seeking and party finance moved away from public finance and donors increasingly focussed on the contacts the government and Frelimo kept, it was a sign that PFM reforms had progressed, at least to some extent. Some of the last areas covered were not surprisingly procurement audits, which were not mandatory and the patrimony (assets register) module in E-SISTAFE. However, even these areas advanced eventually.

The downside of the amount of aid was a high degree of donor dependency. Mozambique failed to integrate these resources, policies, and PFM procedures in the interests of local producers. Moreover, aid incentives were very generous but biased towards, for example, low quality investment in human resources. Some achievements were made, but in a fragile process (Cruz and Mafambissa Reference Cruz and Mafambissa2020). An inclusive and sustainable productive environment remained lacking and difficulties in the donor–recipient relationship discussed by Flentø and Simão (Reference Flentø and Simão2020) intensified. The overwhelmingly benign alliance for social transformation between western donors and the government of Mozambique started cracking in the mid-2000s as both parties changed their ways. One reason was a change of personalities and political line as Guebuza took over from Chissano, but there were strong underlying drivers pushing the divide.

On the donor side the re-securitisation of aid was significant after 9/11 and western military intervention in Afghanistan and Iraq. Aid budgets came under pressure and donors simply changed their yardstick in relation to the underlying principles guiding aid flows. As the unipolar world came to an end, donor countries wanted to spend more aid in fragile states, both post-conflict and the ones they were still bombing, as Afghanistan. Under such circumstances, where western security was on the line, donors would support any government regardless of its democratic and governance credentials. Large flows of harmonized aid to peaceful countries as prescribed in the Paris Declaration on effective aid management became an easy target for revenue collection. These programmes were not very popular with many politicians in donor countries, due to reduced visibility and low possibility of combining aid with commercial efforts. Mozambique was especially vulnerable as all aid was coordinated around GBS, which many donors had joined to get a seat at the dialogue table and not in true pursuit of aid effectiveness. All it took to reduce aid to countries like Mozambique was to interpret compliance with conditionality a little stricter and expose things that western security services had known for a long time about the ruling party’s business finance.

In contrast to the governments in, for example, Kenya, Uganda, and Ethiopia, Frelimo did not comply with the new more or less explicit conditionality. Guebuza strengthened relations with China, did not respond to western concerns over piracy, and maintained friendly relations with private businessmen whom the west believed were drug lords. Realising that conditionality would only get tougher, top Frelimo leaders moved in the first half of the 2000s to strengthen the main party structures and their influence. Frelimo started to assert its dominance over state structures. This meant that policies were increasingly defined by Frelimo and not by foreign interests, which eventually led to tensions with the donor community. It also meant that activities to empower local producers would have to find their way through a centralised system of oligarchs who used their political influence in the executive to dominate the legislature and the judiciary. These oligarchs controlled positions in major public companies and obtained rents from large companies extracting natural resources.

Meanwhile, in the second half of the 2000s, PFM reforms did continue on the previously defined course, backed by aid resources. Mozambique had no star record on human rights and democracy, but in macro-economic management, including PFM, it was considered a very good performer as expressed by IMF in 2007. To keep the GBS and significant sector budget support flowing, the PFM reforms had to continue, even intensify. However, they concentrated on the management and transparency of flows and auditing of an ever-larger part of government entities and not least, the new priority of many donors, tax collection. In the medium term, few of these reforms would seriously reveal or hurt the way top party officials and oligarchs increasingly accumulated wealth. Elite accumulation had much more to do with acquiring land and other assets as well as exclusive licenses to team up with foreign partners in large scale infrastructure, extractive industries, power generation, and telecommunications. Some of these partners were from non-western countries like China and India, others were private corporations from the exact same countries that were providing the bulk of aid to Mozambique.

C The Hope from Gas, but All That Glitters Is Not Gold

With the confirmation of large natural gas reserves in the Rovuma basin, and after a harsh reminder of donor dependency by the GBS-freeze after elections in 2010,Footnote ¹⁵ critical decisions were taken by the government to increase the use of commercial loans to finance a whole series of public investment projects and to provide hidden state guarantees to commercial investments in the Proindicus and MAM companies. Ematum became publicly known because a French shipyard gave interviews to their local press on new orders the company had received from Mozambique. Meanwhile, PFM reforms continued, but they became more tentative and fragile in light of the increase in the public deficit and debt. Roe (Reference Roe2018) characterises the process as one of ‘bust before boom’ and Henstridge and Roe (Reference Henstridge, Roe, Addison and Roe2018) discuss macroeconomic management of natural resources. Expectations of large natural gas revenues put pressure on the PFM system, but the system as such prevailed. When fiscal discipline was blatantly undermined, it was not by embezzlement, gross neglect, or mismanagement, which would be revealed by the system and the checks and balances built in during many years of reform. It took outright fraud committed secretly by the most senior government members to override the system, in an operation where the financial flows out of the system would only ensue several years after the crime had been committed. This is a kind of a hallmark for PFM reforms in Mozambique.

Even a good PFM system cannot discover three secret letters signed by senior ministers in breach of their mandate, but when such letters entail financial flows of public funds it will. Therefore, decisive rupture between the top party/government interests and the PFM reforms agenda only occurred in the first half of 2016 when the ‘hidden debť scandal became known by the public. That happened when the creditors announced they would call in the state guarantees signed by senior ministers in secret letters 3–4 years earlier.

The G19 group of international aid partners stopped all budget support (general and sector) and the IMF suspended its assistance to the government until 2019. When the Nyusi government took over in 2015, it inherited an economic crisis that was further deepened by a series of climatic and political shocks, and the mitigation measures taken came with a heavy cost to the economy. Expectations that economic progress could continue as in previous decades became an illusion and the GDP growth rate declined to very low levels.

To understand why donors reacted so rigorously and relatively united, it is useful to understand how the guarantees issued by the Guebuza administration were conceived by many of Mozambique’s partners.

Having believed they had practically full insight with the governmenťs fiscal policy and budget and after several years of hardening of policy dialogue, the donors discovered that they had not had insight with state finances where it really counted. Especially IMF was disappointed, and logically questions were asked: ‘How could we not know this, have we done a proper job?’ Such questions trigger reactions from people whose career is in the balance and also questions the almost blind confidence donors have in IMF knowing what goes on in an aid dependent economy with large amounts of general budget support.

Originally, donors were not primarily concerned with the real crime of the secret guarantees – the fact that they were not disclosed to and approved by the Mozambican parliament. It was the case of misreporting to the IMF that concerned many donors most. Such misreporting occurs from time to time and although misreporting of this magnitude (almost 10 per cent of the GDP) is rare, it was not unprecedented (the Democratic Republic of Congo, for example, had known worse cases).

However, in Mozambique donors discovered that the government had contracted debt that would consume more funds than the entire G19 budget support would provide over the medium term. The government of Mozambique had secretly mortgaged the donors’ confidence-based financial assistance for years to come. It was all possible because of the future gas revenues, but the extraction of gas was still too far away for those flows to service the debt. Some capital gain taxes from gas companies could probably be collected during the time the debt was to be serviced, but the really big financial flows from gas would only materialise after the loan were repaid.

This timing was probably not the initial intention of the government, which, together with many others, including IMF and donor countries, had very optimistic expectations in relation to how fast the gas fields could be operational. However, the Guebuza administration was sitting on its second and – according to the constitution – last mandate. Whether it was to finance an exit strategy and retirement or to swing constitutional reform in favour of a third mandate, waiting for the funds from the gas exploration was not an option. There was, however, also another reason the loans were urgent.

The loans were not just a bogus operation for personal enrichment of the elite, including foreign bankers and banks with little social conscience. What the government tried was fundamentally to finance its way into the security architecture that would surround the future gas and oil extraction in northern Mozambique. That was a significant additional concern to some donor countries.

By acquiring what looked like a fishing fleet and fast going inspection and patrol boats as well as shipyard service facilities for those vessels, the government of Mozambique, or rather the security sector of Mozambique’s government, could be an important partner in future security operations around the extractives industries in the north. Such a Mozambican structure could arguably be strong enough to choose its foreign partner. That is relatively rare in Africa, where many governments, just like Mozambique’s, lack the resources to do so, among other reasons because security expenditure had been restrained in highly donor-dependent countries.

When a government cannot guarantee the safety of extractive operations in its territory, there are only two other options. One is to let another country do it, which is often highly controversial and directly unthinkable in Mozambique. The other option is privatising the security operations, which is most often the outcome. Contracts for the services (often labelled security and logistics advice and training) are somehow overwhelmingly awarded to a handful of western companies.

All along, the PFM reform agenda continued to be pursued. There was a reform schedule to improve various procedures, correct more recent weaknesses (restructure MEF bodies, improve debt management, and public sector risk assessment), and roll out the system to all public entities and all spatial levels. Moreover, there were previously committed resources available, whether internal or external, and a sense of credibility by advancing with the reforms. In sum, the continuation of the reforms represented a professional commitment from those officials directly in charge and all those entities, both local and international partners, dedicated to the improvement of mechanisms and management of public financial resources.

Nevertheless, internal, and external trust in the state became very low. A former Minister of Finance and member of the Mozambican parliament has at the time of writing been in jail since 2018 in South Africa (Hanlon Reference Hanlon2019).Footnote ¹⁶ Moreover, in the beginning of 2019, a former Treasury National Director and Deputy Minister of Finance was dismissed due to her links to the hidden debt. These and other events contributed significantly to demoralising civil servants working directly on improving the PFM system. More broadly, the ‘hidden debť scandal and its aftermath contributed to the decline in moral authority of high-ranking state officials. The quality of public services deteriorated and opportunities for corruption increased and became endemic. The continuing economic and political shocks that hit the country from 2015 onwards, including the insurgency in Cabo Delgado since 2017, have moved policy focus away from the more detailed needs of the PFM agenda, which has become less of a priority.

What we described above is reflected in a series of Public Expenditure and Financial Accountability (PEFA) reports.Footnote ¹⁷ Overall progress in PFM indicators has deteriorated in recent years reflecting the reduction of available resources and the need to respond to urgent priorities, such as the natural disasters and since early 2020 COVID-19. Moreover, the enabling environment for PFM reform clearly deteriorated due to the abrupt stop to Budget Support in 2016.

A The Role of Donors

The story of PFM in Mozambique shows that donor prompted reforms will only work in the long run, if they are truly owned and exercised by the host government. Even the best reforms from the technical point of view will not solve the fundamental challenges if they are imposed in exchange for amounts of aid so large that the government cannot afford to decline. As alluded to by Bourguignon and Gunning (Reference Bourguignon and Gunning2016), aid can be for pure finance or in exchange for reforms. In practice they overlap but conceptually they are very different. When aid is used to promote reforms, the fundamental difference in objectives between the government and donors is the key reason for conditionality. If there is no disagreement, conditionality is pointless, in the other extreme aid acts as a bribe.

The donor community applying GBS shared a belief that well-coordinated GBS could be used as a lever for almost any reform and starvation of the security sector. This has proven unrealistic in relation to reforms on governance and democracy that would entail severe risks to the holding power of sitting governments. The PFM reforms in Mozambique basically avoided this and were allowed to progress and improve fiscal management for more than two decades. When Mozambique’s government no longer could pursue key priorities due to the PFM-reforms and donor dependency, they circumnavigated the PFM system by secret, illegal action.

This was possible because reforms were not comprehensive across sectors. Many laws and decrees had been passed, as well as procurement, installation, and training in new systems, in what had been a true and continuous upgrade of PFM during more than two decades. It looked good, also to donors, which believed they had assisted the reform-friendly parts of government against the hard-lines in Frelimo by providing generous amounts of aid, including GBS. But the fundamental anchorage was missing. Donors were to a large extent deceived by a phenomenon referred to by biologists as isomorphic mimicry. When a non-venomous snake looks like a very poisonous relative, it can help frighten enemies and predators. Yet it cannot hunt like a venomous snake if it has no poison. PFM reforms in Mozambique can be characterised as isomorphic mimicry. The reforms did help in directing the economy and provided the basis for critical scrutiny and exposure of mismanagement and embezzlement. But without an independent legal system – state prosecutor and courts – to back it, the fundamental challenges could not be addressed. The snake had no venom.

B The Separation of Powers and the Rule of Law

The necessary checks and balances for a democratic system to work were formally introduced in the 1990 Constitution of Mozambique but became gradually dysfunctional from 2005 onwards. In a presidential system, the role of the president in the government is key. In addition, in Mozambique, the president of the Frelimo party has also been the President of the Republic. The president appoints and dismisses the presidents of the Supreme Court, the Constitutional Council, the TA (external auditor), and the attorney general. As such, the incumbent holds significant power. Moreover, in parliament, Frelimo has had an absolute majority since 1994, and party members in parliament respond to centralised party instructions. In such a tightly controlled environment it is not surprising that when members of the elite decided to increase the demand for non-concessional credit, there was no effective mechanism to prevent systematic breach of the law and illegal actions.

The Mozambican parliament has a good track record in producing legislation and organising the legislatures, and the many PFM reforms have contributed significantly to this being possible. In the case of the hidden debt parliament could not exercise full oversight in the management of the budget, because senior ministers committed fraud secretly. In 2013–14, the government in effect bypassed parliament illegally and contracted direct liabilities by issuing state guarantees for debt. Subsequently, in 2016, when this information became public, parliament could have, but did not act upon it. There were no immediate or direct consequences for those responsible for violating the constitution and the budget laws. The lack of action in response to the infringement of key laws implies that the majority of the members of the parliament and the Budget Committee were either co-opted politically on this specific issue and did not aim to enforce the law, or lacked the effective power to do so, just as it was the case for the judiciary (Orre and Rønning Reference Orre and Rønning2017).

When it became clear to donors that no amount of withheld aid would make the Mozambican judiciary perform in this case, the US government acted to encourage the Mozambican judiciary to react by issuing international arrest mandates for the officials involved. Due to the risk of those arrested to enter plea-bargains with the US courts, the Mozambican state prosecutor acted in February 2018. Once legal action was finally taken against the culprits, the way was paved for declaring the guarantees and debt illegal. In May 2020, the Constitutional Court ruled on the secret debts, considering them ‘null and void’ (Hanlon Reference Hanlon2020).

Similarly to the first historical phase in post-independence Mozambique, when political power was overly concentrated and the system of checks and balances was not operational, recent economic policy decisions can have large and lasting effects on the economic performance of the country and on the welfare of most citizens. In sum, Mozambique is a case where the party in power has held excessive political control, confronted by a weak opposition, a parliament and a judiciary that has been either unable or unwilling to act against obvious attempts to undermine the otherwise respectable progress during the 1990s and 2000s in the PFM system. Socio-economic consequences have been dire.

C Decentralisation

Principles of decentralisation were embedded in the legislation at the very beginning of the democratisation era in Mozambique. The new Constitution of the Republic of Mozambique in 1990, the Local Public Administration Reform Programme (PROL) in 1991, and the Constitution’s Amendment through the Law 9/1996 allowed for the creation of municipalities composed of an executive organ, the Municipal Council, and a legislative organ, the Municipal Assembly (Fernandes Reference Fernandes2007; Simione et al. Reference Simione, Matos and Ckagnazaroff2018).

The decentralisation process did lead to greater autonomy in urban areas, thus giving more power to the local authorities responsible for improving the livelihoods of their citizens (Weimer Reference Weimer2012). The gradual increase in the number of municipalities has meant that decentralisation was understood as a process of expansion of urban structures to rural areas. However, the transformation of the rural areas into municipalities, as part of the decentralisation process, has been seriously lacking. There were only 33 cities and villages considered to be municipalities in 1997, 43 in 2008, and 53 in 2013 (Brito Reference Brito2019), and this number has remained unchanged since. Rural areas that are not included in municipalities are governed by the government branches at provincial, district, administrative post, and locality level, meaning that Mozambique has effectively a highly centralised organisational system.

Since 1993, a pilot decentralisedFootnote ¹⁸ planning and budgeting exercise was carried out through the Decentralized Planning and Finance Programme (PPFD), which was replicated in other provinces. Capacity building, including training, application of new procedures, and investment in equipment, has been provided for more than twenty years in this field. The results are positive, but the process is far from complete, and it is necessary to further develop the model of decentralisation suitable to the country.

A further dimension of the decentralisation process was the establishment of provincial assemblies in 2009 and the election of provincial governors in 2019 (Brito Reference Brito2019). These are reforms of a primarily political nature, but their viability relies on a well-performing PFM system (PoM 2018). Political, military, and civil society activities have been leading to decentralisation reforms, including the PFM system. However, the success of these reforms will depend on the allocation of quite large amounts of financial resources and current organisations in the state apparatus, elected assemblies, and political positions have been under severe budget constraints since 2015.

PFM reforms should help ensure that provinces and districts are entitled to decide over and manage public funds. Law 8/2003 of State Local Authorities and Decree 11/2005 established legal principles and norms for the decentralised levels of government (PoM 2003; GoM 2005a). This legislation allowed for greater budgetary autonomy at the provincial and district levels and has been implemented since then. However, the unequal distribution of the budget in per capita terms among provinces remains a core challenge (UNICEF 2017).

Political instability and military conflict are partially explained by the insufficient degree of decentralisation. This has been happening possibly due to a perceived risk by the party in power of losing political power at the local level and potentially at the national level. The political bottlenecks in the decentralisation process, the per capita inequalities in budget allocations, and the political interference in the budget execution criteria at local levels, in particular at district level, distorts the proper functioning of the PFM system.