Background

Human resources for health (HRH) represent a critical element for the functioning of health systems. Poor availability and management of human resources have been recognised as key health system barriers, and despite the efforts, many low-income countries continue to face acute HRH challenges [1, 2]. These challenges are compounded in conflict-affected settings where health needs are exacerbated as a result of the conflict and the health workforce is often decimated by either death or flight due to violence [3, 4]. Consequently, the challenge of attracting and recruiting the right persons into the health workforce, according to needs, becomes more acute and relevant [5,6,7]. Some studies in low and middle-income countries (LMICs) have focused on Human Resource Management (HRM) [8,9,10,11,12,13,14] and highlighted how recruitment affects retention in remote areas [15,16,17,18,19]. Research has also looked at the specific HRM challenges in fragile and conflict-affected countries [5,6,7, 20,21,22,23,24,25,26,27], with a focus on health worker retention and (financial and non-financial) incentives.

This research adds to that literature by focusing in particular on the role that recruitment policies and practices play in rebuilding an appropriate health workforce after conflict and following the trajectory and patterns over the subsequent years. Our case study was conducted in Timor-Leste starting from the immediate post-conflict period up to 2018 (Table 1 and Fig. 1), which gives a retrospective view covering a period of almost 20 years. This paper focuses on policies and policy-making processes at central level related to health workforce recruitment and presents a political economy analysis about how and why both official and informal practices developed, as well as the drivers, challenges and blockages at different stages.

Table 1 History and context of Timor-Leste
Full size table
Fig. 1
figure 1

Oil and gas revenues in Timor-Leste (2009–2015) (US$ million). Source: [43]

Full size image

Methods

This study is part of a broader research covering HRH recruitment and deployment in Timor-Leste and exploring both the central and sub-national debates. In the present paper, we focus exclusively on the central-level perspective. The research adopts a retrospective, qualitative case study design [28], drawing from documentary review and key informant interviews. Policy analysis was conducted, with elements of political economy analysis which allowed the identification of how elements of the agency (actors, agendas, power relations) and structure (socio economic conditions, historical legacies, formal and informal institutions, cultural norms) contributed to drive the policy trajectory [29].

Data collection

A document search was carried out to identify published literature and documents relevant to HRH recruitment policies and regulatory frameworks in Timor-Leste between 1999 and 2017. Internet searches were carried out using Google, PubMed, HRH Journal and Government websites (e.g. Official Gazette, Public Service Commission, UNTAET legislation archives). Additionally, key informants were asked to suggest relevant documentation, and the contextual knowledge of the team in country also allowed retrieving further documents. In total, 76 documents were reviewed (a summary is provided in Additional file 1).

Key informants were purposefully selected to include a wide range of actors involved in HRH policy-making over the study period. Twenty key informants were interviewed (Table 2) by AAG and JM between March and May 2018, either face-to-face or remotely, in Tetum, English, Spanish or Portuguese depending on the location and the language preference of the respondent. A standard topic guide was used flexibly to adapt to each respondent’s knowledge and elicit a retrospective enquiry about the main HRH challenges with a focus on recruitment, policy solutions to HRH issues, changes in policies and practices, drivers of change, actors and agendas (Additional file 2).

Table 2 Summary of key informants by type and gender
Full size table

Data analysis

Information was extracted from the documents using an Excel-based template including an analytical thematic framework developed for this purpose (Additional file 3). Based on this, an initial draft of a ‘policy timeline’ was prepared which described in chronological order the key events and changes that had happened. The documentary analysis also helped to identify gaps in the information available, so that these could be addressed during the key informant interviews.

Key informant interviews (KII) were recorded, transcribed and translated into English (where needed). Verbatim transcripts were manually analysed, using a series of pre-defined descriptive themes (Additional file 3), but also focusing on emerging cross-cutting, analytical themes, more apt to capture the political economy issues and the dynamics which shaped the decision-making processes [30]. Data triangulation was carried out between different sources of data, comparing the narratives of different key informants, but also triangulating information from informants and documents. Based on the KIIs’ analysis, the policy timeline was revised, updated and enriched and key elements, themes and patterns were teased out and discussed among all authors before the final drafting of the article.

Ethical approval was obtained from the Liverpool School of Tropical Medicine and from Timor-Leste National Health Institute.

Results

In this section, we present the main findings of our analysis following a chronological order to show how events unfolded overtime. We also highlight the key turning points and the main drivers and dynamics that shaped the policy-making processes.

Immediate aftermath of Indonesia’s withdrawal and the transitional period (1999–2002)

In September 1999, with the violent withdrawal of Indonesia, the health system, like much of the rest of the social, economic and organisational structure and the infrastructure of Timor-Leste, had been destroyed [31]. In terms of HRH, the damages were profound. Most of the 135 doctors working in Timor-Leste under the Indonesian government were non-Timorese and left; only about 26 doctors remained [32]. Other critical staffing issues included the shortage of midwives, technicians and health managers; the rural-urban imbalances which were further exacerbated by the insecurity following the 1999 violence; and the excessive number of nurses inherited from the Indonesian system [33].

Faith-based and non-governmental organisations (NGOs) were the first to be able to provide basic health services [31, 34], and to do so, they brought in expatriate staff, but also recruited and paid local health personnel available in the districts [35]. Key informants reported that recruitment criteria were rather loose as it was not always possible to find formally qualified staff to fill the position and vacancy advertisement and recruitment were often based on word-of-mouth.

We did not do the recruitment per merit. We just indicated people. […] I called them to the health centre and proposed them to the expatriates of the NGOs. If they agree, we’d recruit them. (MoH KI)

While much of the technical and financial support for the health sector (including HRH) was provided by external partners [36], the Timorese health leaders of the nascent Ministry of Health (MoH) were keen to strengthen their role and increase the legitimacy of the state and the credibility of the emerging government through the so-called timorisation process [32, 33]. In line with this state-building aspiration, it was important for them to build a health workforce that would distance itself from the Indonesian one, considered inefficient. Key informants recalled:

The administrators took the line that it wasn’t just matter of putting people back into the positions they were in and paying them, but it was creating the whole civil service. […] They were very much trying to get a sense of unity [through the civil service]. (international KI)

This overhaul was to be achieved in three ways. Firstly, it was imperative to rationalise the number of health workers (nurses, in particular) employed in the public sector which was inflated for political and patronage reasons under the Indonesian occupation. Secondly, there was a desire to reduce or eliminate corruption, collusion and nepotism (often referred to using the Indonesian acronym, KKN) that was a key feature of the previous system and introduce a meritocratic approach. Finally, it was important to create a health workforce loyal to the new country, excluding those who had supported the pro-Indonesia militia during the independence struggle (KIIs). The aim was to ensure equitable access to quality service for the entire population of the newly formed state.

Our leaders wanted to build a public administration that was not a replica of the previous system. So, when we defined the regulations to form our public service, we did not look at the Indonesian model which was disproportionate. We started from zero. (national KI)

These objectives laid the foundation of the first civil service recruitment which started in mid-2001 (UNTAET Regulation 2000/03). The final estimate of the health staff needed was of 1 242 as agreed between UNTAET, MoH and the Civil Service and Public Employment Office (CISPE). Informants reported that this number, about 47% of the 2 632 in place during the occupation, was mainly dictated by the budget available, which was limited and dependant on international partners [37]. Fifty job descriptions were developed, vacancies opened, and recruitment of health staff began. In order to ensure staff retention in remote areas, vacancies were initially tied to specific locations rather than a central pool from which new recruits would subsequently be deployed (an approach which was later adopted). According to one key informant, in line with the aims of the recruitment, selection criteria included professional training, place of birth and perceived loyalty to the new state, against the Indonesian regime.

The first criteria was that the person had to be trained as a nurse or healthcare professional. The second, we had to look for people that were from that district to facilitate them going there […]. Then the third criteria that we considered was that it had to be people that in 1999 didn’t run away to Indonesia. These people are a priority. (MoH KI)

However, a number of issues hampered the recruitment process. The shortage of doctors remained a challenge and terms and conditions provided by NGOs were much more favourable so that many stayed in that sector. Additionally, strikes and protests broke out due to accusations of favouritism and mismanagement. One informant recalled:

The Bishop came down to calm the situation at the hospital and to listen to the demands of the demonstrators. They were saying that those who had fought for independence had not passed the recruitment. So, I went with [xxx] and we modified the recruitment. Those who had stayed would have work first, and those who went to Indonesia, they have to work last. (MoH KI)

In 2001, 724 workers were recruited (Table 3). The remaining vacancies were filled by international staff recruited by NGOs or by the MoH with Australian Government’s funds.

Table 3 Public workforce situation (October 2001)
Full size table

Recruitment was reported to be slow and haphazard leading to poor morale and loss of trust in the administration by those involved in the process [32]. Their analysis lucidly highlights the trade-off between incompatible ‘technical’ and ‘political’ objectives (Table 4) [32]. However, despite the salience of the HRH challenges identified then, the immediate health sector reconstruction in Timor-Leste and the model of post-conflict health system rehabilitation has generally been judged a success [33, 38].

Table 4 Competing goals for the recruitment of civil servants for the health sector
Full size table

A ‘game changer’: the involvement of the Cuban Medical Brigade (2003–2005)

The idea of requesting Cuban support to address staff shortages had been discussed since 2001 [39]. In February 2003, this option was adopted at top political level when an agreement was brokered between the then President Xanana Gusmao, José Ramos Horta (Minister of Foreign Affairs) and Fidel Castro for the provision of Cuban doctors as well as the training of Timorese students. The Cuban involvement was described by one key informant as ‘a game changer’ for the health sector. In December 2004, the engagement was further detailed as the training of 1 000 doctors (calculated as roughly 1 doctor per 1 000 population) at the Latin American Medical School (LAMS) in CubaFootnote 1 and the deployment of 300 Cuban doctors to provide medical care across the country as well as to supervise and further train the medical students and new doctors [39].

The use of subsidised training effectively shifted the focus from post-graduation recruitment procedures to pre-service education scholarship awards, as those trained in Cuba were expected to return to serve in Timor-Leste and were automatically absorbed in the public health sector.Footnote 2 While in recent years, the absorption of medical graduates into public service is becoming problematic (see below), at the time, it was clear that scholarship assignment became a synonym of hiring.

Instead of recruiting, the MoH is appointing through training. (national KI)

The selection of students to access Cuban medical training was undertaken initially by the MoH, based on academic results in secondary school, a written exam, an interview and a physical examination (KIIs). Gender balance was also sought in the selection, and quotas were introduced based on the geographical origin of the candidates,Footnote 3 with the intention that they would be deployed in their districts of origin ensuring an appropriate coverage across districts, increasing retention and health worker acceptability for local communities. Although the geographical criterion initially created difficulties due to the low skills of those coming from remote areas, these were overcome in subsequent years.

By mid-2018, the Cuban-supported programme had graduated 934 Timorese medical doctors (KII). Based on the MoH database, this rapid scale-up of the medical workforce has increased the density of doctors per population from 0.03 to 0.71 per 1 000 in 2000 and 2017 respectively, among the highest in South-East Asia (Fig. 2). In 2017, the gender distribution of the medical workforce was almost equal with a male to female (M:F) ratio of 1.01 although the balance was in favour of males at specialist level (M:F ratio of 2.6).

Fig. 2
figure 2

Density of doctors per 1 000 population in the South-East Asia Region. Source: WHO Global Health Observatory data repository (http://apps.who.int/gho/data/view.main.92100)

Full size image

Relapse into violence, stabilisation and expansion period (2006–2012)

The relapse into violence in 2006 and the tense political environment affected the state’s legitimacy [40] and led to a degradation of accountability and transparency with increased perception of corruption and nepotism [41]. This probably influenced recruitment practices for the health workforce. For example, it is possible that the settlement of the dispute between government and ex-FALANTIL soldiers led to the provision of privileges for the families of veterans of the armed struggle. These privileges included the possibility of obtaining scholarships for higher education bypassing the entry requirements and examinations (Decree Law 8/2009), which opened the system to opportunities for discretion and patronage.

From 2011, the responsibility for selecting students into higher education for all health professions was moved to the Ministry of Education (MoEd) (KII). Key informants highlighted the alarming absence of functioning mechanisms for inter-sectoral coordination, which not only left room for discretion and abuse in decision-making, but also had important consequences for the development of the health workforce. Indeed, the fact that decisions about access to training were moved outside the MoH and some groups were exempted from the standard selection procedures had (and continues to have) critical consequences in terms of quality as well as of numbers and profiles of the future health workforce and may determine a mismatch between intake and needs which the MoH has been left to deal with.

The MoH is not involved in any decisions on the number of students. […] There are no mechanisms for coordination between the MoH and the MoEd. This is a little bit very sad. […] If we recruit too much it will be a disaster for our country. (MoH KI)

The tensions between different national institutions also reveal that, from the aftermath of the independence, the decision-making arena has become more crowded. As a consequence, a number of emerging competing agendas and interests made coordination and alignment more difficult. New actors involved in HRH recruitment and deployment practices included not only the MoH, the MoEd and National University of Timor-Leste (Universidade Nacional Timor Lorosa’e—UNTL), but also the Ministry of Finance, the Public Service Commission (PSC), the Human Capital Development Fund (HCDF) and, more recently with the attempt to decentralise the deployment process, the Directors of Municipalities.

As the number of actors involved in HRH decision-making increased, so did the official legislation applicable to the civil service in general and to health workers specifically (Table 5). Importantly, key informants stressed that these regulations were often overlooked or weakly implemented. Indeed, in the increasingly fragmented and instable political environment with constant changes in governments, policies were often blocked at approval or before implementation stage. In other cases, funds are lacking to support implementation.

Table 5 Key official laws and regulations on HRH recruitment
Full size table

The implementation of the policies is not consistent. One government comes in with their own policy, then suddenly there’s another government with a different policy and that cause the inconsistencies. (MoH KI)

The increasing number of actors involved in HRH processes, their conflicting agendas, the absence or non-respect of formal regulations and the weak management systems all contribute to the weakening of formal institutions and the emergence of informal practices. Over time, the recruitment process appears to be reverting to the previous practices of patronage and discretion and past hierarchies. This was noted by the key informants.

In relation to recruitment for the MoH, we undergo procedures that are not necessarily based upon the principles of meritocracy. […] We observe the issue of corruption, collusion and nepotism, the family approach. Why this is happening? Because, apart from being a post-conflict country, there are also demands for livelihoods. At the same time, we have significant gaps in regulations and procedures that define the recruitment process and therefore it was still dominated by a situation called ‘calling each other’ (politika bolu malu). (national KI)

HRH reforms under fiscal constraints (2013–2018)

The HRH challenges highlighted above have been made starker by the fiscal constraints, which increased in 2012 when oil revenues dwindled and has worsen since 2015 when the reserves in the Petroleum Fund begun to diminish [42, 43]. Yet a World Bank study [44] reported that the health wage bill increased by 344% between 2008 and 2014, mainly due to the absorption of the newly graduated doctors into the public sector (Table 6), despite the fact that doctors’ salaries are not substantially higher than those of other health professionals [45].

Table 6 Number of staff by cadre in the public health workforce
Full size table

In this context, the absorption into the public sector of the growing number of health workers became a pressing problem. Although estimates vary (from 200 to 600), a high proportion of the pool of qualified doctors and other professionals is currently unemployed, due to lack of funds from the Ministry of Finance to recruit them into the civil service. Worryingly, pressure will be mounting over the next few years as the sustained production of health workers will continue and even increased, based on politically pressured student intake. Indeed, beyond the privileges accorded to veterans’ families described above, the introduction in 2016 of a ‘special regime’ to access higher education, following a simpler selection process without technical tests, for families of military and police officers, diplomats, members of Parliament, journalists, elite athletes and students from international secondary schools [46] contributed to this problem. While in principle, only 10% of the annual intake should be recruited through this modality, key informants noted that the system is abused because of the (politically pressured) acceptance of a high number of students through the special regime. For instance, in 2018 more than 65% of students enrolled in Nursing studies, 60% of those in Midwifery and 35% in Medicine went through the ‘special regime’ (Table 7). While the Cuban Medical Brigade (CMB) deliberately kept out of the politics and simply scaled up their training capacity, the choice, as one respondent explained, has obvious long-term consequences for the quality and distribution of the workforce:

Table 7 Student intake in the Faculty of Medicine and Health Sciences of UNTL (2004–2018)
Full size table

We need to fix the special regime because the veterans’ children have low marks, but they all choose Medicine. […] If they don’t have a good understanding [of Medicine] then they might give the wrong medication or make mistakes, and this is a problem. (national KI)

As an interim measure to address the absorption issue, a dispatch from the Minister of Health (07/2018/I/MS) was issued in January 2018 to recruit 320 recently graduated health workers as interns (regimen de internato) with a standard stipend regardless of professional level of $200 (in rural areas) and $150 (in urban locations) for a period of 6 months to be extended if the need persists and the performance of the intern is satisfactory. However, as KIs said, the success of this programme in reducing the tensions is yet to be seen, and there is a widespread perception that the government has been and continues to waste money on the training of medical doctors rather than using data and evidence on the real needs.

The contrast of the current situation with the immediate post-conflict phase could not be starker. As one respondent put it:

At that time, we didn’t have the people, but the money was there. Now it is probable that we have too many people but not the money. (MoH KI)

Discussion

Our analysis has some limitations. First, the long period of time covered, spanning over almost 20 years, created challenges during data collection, both in terms of locating and accessing documents that may have been lost, as well as contacting key informants, in particular international actors who have left Timor-Leste. Data were also difficult to retrieve due to the weak information systems. Additionally, many interviewees struggled with recall bias. However, the fact that respondents were open about it and distinguished between clear and only vague memories helped us make sense of their accounts. In a sense, even the mental selection between remembered/forgotten could be interpreted as an indicator of the key, most debated and contested areas of decision-making. In line with this, in order to elicit views on issues clearly remembered and to reduce recall bias, the topic guide we developed focused on the most relevant changes or challenges which respondents were asked to reflect on. We also carefully applied triangulation between different key informants as well as with documents and data to ensure the accuracy of the information and the relevance of the analysis.

Throughout the analysis, we have been explicitly reflective on our positionality with respect to the study subject [47]. Some authors (AAG, JM and SP) are ‘insiders’ to HRH policy-making in Timor-Leste and, although in different roles, have been directly involved in the policy changes described. This has advantages in terms of accessing information and actors as well as bringing participant observations to the analysis, but it may reduce objectivity [48]. The presence of ‘outsiders’ among the authors allowed some distance from the study subject, while at the same time data interpretation could be collectively and iteratively reviewed.

The study findings provide critical insights to enhance our understanding of HRH policy development from a policy and political economy perspective, in a post-conflict setting, but also following the developments over the longer period. The analysis allows the identification of the main phases of the policy trajectory during which different drivers played a significant role in shaping policy outcomes. In terms of actors, there was a shift over time as the relevance of external players faded in favour of that of internal actors and institutions. As in other settings during the immediate post-conflict phase [3, 49,50,51], Western donors and multilateral organisations were essential and influential in the aftermath of the violence to provide financial and technical support to the nascent Republic, including supporting HRH systems. Later, Cuba became a prominent player, and despite their professed disengagement from ‘politics’, their views and approaches (e.g. gender and geographical balance of students selected) undoubtedly influenced policy-making. More recently, the trajectory in Timor-Leste took a different direction to many post-conflict countries, as national actors have emerged to play a more prominent role and their number and influence have increased in an increasingly fragmented political landscape [52]. It is important to note that the shift in actors’ roles is closely related to and driven by structural changes (specific to Timor-Leste’s economic history) and in particular the move from aid dependency in the first phases of the reconstruction, to financial independence due to oil revenues and, more recently, budget constraints related to the diminishing revenues. Over time, the structural features remained informed by the formal and informal institutions (i.e., the rules and norms) which are prevalent in the country and underlie the overall political and economic context.

Agency and structure elements and their interplay influenced the policy processes and outcomes. Initially, the focus was on health workforce recruitment as is often the case [27], which was seen as fundamental not only for its contribution to the health sector but also in relation to the state-building efforts [6]. As in other settings [53], a key emerging issue was the challenge of achieving a balance between the ‘what’ (technical) with the ‘how’ (political) [52] and reconciliation of a technocratic recruitment process based on rules and merit with political, cultural, relational norms and practices.

It appears that the recruitment system created in the aftermath of independence did consider both technical and political aspects and, perhaps because of this, was generally effective and survived over time. The relative success of this first round of HRH reforms could also be linked to specific individuals, their skills and leadership, both among external and national actors. Local elites were quick to organise themselves with enthusiasm and focus [32], also because of their state-building aspiration through HRH recruitment [6], and there was a substantial alignment in their agendas with that of the technical assistants, as also noted by others [36]. For example, the need felt by national decision-makers to distance themselves from Indonesian rule and ‘KKN’ was in line with the technical and meritocratic approach of external advisers. Additionally, the fact that initially there were few actors in the decision-making arena limited the presence of competing agendas and disagreements. However, while this led to the re-establishment of a HRH recruitment system, the fact that the focus of our analysis rests predominantly on recruitment processes and doctors (rather than on other HR management issues or on other health professionals) highlights how these elements were priorities for policy-makers at central level, while other cadres and issues (such as, deployment) remained under-regulated and dealt with at decentralised level, with the lack of overall strategy and clear official policies to enact, as documented also elsewhere [11, 12, 54]. Additionally, policy-makers paid little or no attention to health workers’ job preferences [45, 55] and decision-making remained a top-down exercise.

Later, while the technical foundations of the recruitment system remained in place and were actually strengthened by institutions such as the PSC, the shift to ‘appointment through training’, due to the Cuban-supported training and to the virtually automatic absorption of newly graduated staff, meant a shift of power from the MoH to new actors (such as, MoEd and UNTL) with diverging interests not related to the long-term needs of the health sector, and increased the possibility of rule-bypassing, discretion and patronage.

The tension between competing agendas may have been initially reduced or resolved thanks to the fiscal expansion and to the presence of the CMB and their ‘substitution’ role [56, 57]. However, with decreasing oil revenues, the CMB potentially phasing out and increased political fragmentation, the competing interests are more difficult to reconcile and the tensions appears starker and riskier in the long term, leading to a low-quality workforce, inappropriate skill-mix, geographic maldistribution and financial unsustainability [58]. In the future, it will be essential that policy processes in Timor-Leste lead to more robust, sustainable and better implemented strategies on HRH training and recruitment.

Conclusions

Our research aimed to provide a long retrospective account of the development of the systems to recruit health workers in post-conflict Timor-Leste. In this study, we explored the policies in place at central level, the main phases and key drivers of their evolution. Our findings point to patterns and elements both of the agency and structure in shaping the policy-making processes and outcomes.

While the influence of international actors was a prominent feature of the initial phase after the conflict, in Timor-Leste, there was an important change as national actors increased in numbers and relevance later on. This shift led to a fragmented institutional landscape with diverging agendas and lack of inter-sectoral coordination, to the detriment of the long-term strategic development of the health workforce and the health sector. Furthermore, a key issue that cuts across all phases is the difficulty in reconciling the technocratic elements of the reforms with the social, cultural and political aspects. Finally, the research component discussed here looks exclusively at processes and dynamics at central level, and finds that they have focused on regulating HRH recruitment and production, which are the most visible aspects for those operating centrally. Further research on issues such as deployment and transfer, motivation, promotions and career paths (unregulated areas at central level) as well as on how implementation practices differ, modify or bypass central regulations, from the perspective of decentralised actors will be important to shed further light on the topic and complement the central-level perspective.