Why is there often a gulf between what policy-makers intend and what is implemented?
Illustrate your arguments with two examples with which you are familiar.
The gap between policy intention and implementation has attracted great research interest (Matland, 1995), yet still, many policies end up in failure (O’Toole, 1986). Different reasons can be advanced for this disconnect. These range from insufficient resources to bureaucratic discretion, ambiguity to poor framing of policies and poor policy instruments (Colebatch, 2009).
The two main frameworks that shape the implementation field are the ‘top-down’ and ‘bottom up’ theories. Top-down theorists view policymakers as central, and policy implementation as manipulable and successful implementation as meeting the of the policy (Howlett, Ramesh and Perl, 2009). Conversely, bottom-uppers view policy implementation as decentralised (Barrett – 2004) and therefore emphasise on target groups and service deliverers. They view implementation success as the adaptation of street –level bureaucrats to achieve policy goals (Howlett, Ramesh and Perl, 2009).
Success and failure are ambiguous, yet failure is inextricably linked to policy implementation. I make a distinction between non-implementation as the total failure of a policy and unsuccessful implementation as the partial failure. This is in line with Colebatch (2009) definition of successful policy implementation as achieving the goals of a policy.
In this essay, I will do a theoretical and empirical analysis of two policies whose implementation failure exemplifies points to the gap between intention and implementation. The first section will focus on Kenya’s Free Primary Education (FPE) and the second section will analyse Kenya’s Anti-Corruption policy. The study of these two cases will rely on two different theoretical frameworks, namely groupthink and ambiguity-conflict theories.
Kenya’s Free Primary Education (FPE) Policy
The Free Primary Education (FPE) policy was introduced in 2003, by the new coalition government National Alliance Rainbow Coalition (NARC). This policy aim aimed to reverse the hitherto declining primary school enrolment in Kenya, by allowing school-age children from all socio-economic backgrounds to access basic primary education (Uwezo, 2015; Kaimenyi, 2016).
Serious problems however hobbled the implementation of this policy. (UNESCO, 2005; Oketch and Ngware, 2010). To begin with, the government wasn’t well prepared to implement the FPE policy which was introduced in the middle of the financial year 2002/2003, thus wasn’t budgeted for, but had to be rapidly implemented to fulfill election promises (Somerset, 2009; Oketch and Ngware, 2010; Abuya et al., 2015). The new school year was starting barely a month after the elections, and the introduction of this policy caught all stakeholders unawares. The schools were left to their own devices to implement this policy (Somerset, 2009; Abuya et al., 2015).
Not surprisingly, in a country where a many children were out of school, the response to the FPE policy was massive (UNESCO, 2005). Primary school enrollment rose from about 6 million in 2002 to about 7.2 million in 2003. This accounted for 104% of the gross enrollment rate (Norrag, 2003; Riddel 2003; Abuya, Oketch and Musyoka, 2013; Abuya et al 2015; Uwezo, 2015). Unfortunately, Kenya had previously underfunded the education sector and the large influx overwhelmed the already inadequate existing resources (UNESCO, 2005).
Classrooms became congested – some up to double their capacity – and learning facilities like desks and textbooks were insufficient (UNESCO, 2005). Teacher-pupil ratio drastically reduced and disparities in class size soared. Moreover, teachers’ work burden increased, yet their salaries remained stagnant (Uwezo, 2015), leaving them unmotivated. Due to these (and other) challenges, school performance and education quality quickly deteriorated (Uwezo, 2015), killing the confidence in public school system. This negated the initial gains of the FPE policy as enrollments fell and dropouts increased (Oketch and Ngware, 2010; Abuya et al 2015; Uwezo, 2015).
Despite the ‘free’ education, parents pulled their children from public schools to private schools, whose enrollment nearly tripled (Riddel, 2003; Oketch and Ngware, 2010; Uwezo, 2015; Kaimenyi, 2016). This switch in education choice also touched poor households as about 60% of families living in urban slums opted for low-cost private schools rather than the public schools (Abuya et al., 2015). Kaimenyi (2016) rightly notes that it is both unfortunate and ironical, that a policy whose intention was to reduce the demand for private school, actually led to the opposite. In fact nearly 59% of pupils who had enrolled for grade one in 2003 did not reach final grade (Abuya, Oketch and Musyoka, 2013).
I argue therefore, that the FPE policy failed right at the decision-making stage, as it took a top-down approach. Some of the key reasons that can be advanced for the failure of the FPE policy includes: the lack of initial stakeholder involvement (UNESCO, 2005), hurried implementation before necessary infrastructure was in place (Somerset, 2009), low quality education which drove children to private schools (Riddel, 2003; Uwezo, 2015; Kaimenyi, 2016) and high costs involved in implementing this policy (Riddel 2003).
Theory and Analysis
Despite being an incredibly important pro-poor policy, FPE’s failure points to defective policymaking in the policy design stage. Many theories can be used to explain the failure of the FPE policy. In my analysis, I argue that the formulation of the FPE policy was marred by groupthink, leading to the disconnect between policymakers’ intention and the resulting implementation.. My argument is hinged on Irwin L. Janis (1982) groupthink theory.
This essay uses Janis (1982) definition of groupthink as ‘a mode of thinking that people engage in when they are deeply involved in a cohesive in-group, when the members’ striving for unanimity override their motivation to realistically appraise alternative courses of action’ (p9). The theory predicts that when a group is highly cohesive and interacts with other antecedent like structural faults and/or provocative situational contexts, this increases the likelihood of groupthink tendency. The groupthink is then observed as a consequence of the symptoms of groupthink. Groups that display symptoms of groupthink have a high of making defective decisions, which end in failure.
Janis (1982) emphasises the three antecedents which underpin the groupthink theory. These are group cohesiveness; organisational structural faults; and provocative situational contexts. All these three main antecedents were present in the FPE policymaking stage. The NARC coalition party was a cohesive and homogenous group, formed to the major objective of the general elections. The opposition parties understood that it was nearly impossible to win against the incumbent President’s party unless they joined up in a coalition (Barkan, 2004). This group was led by five key politicians aptly named the “dream team”, which worked together to deliver a landslide win in the general elections. Parsons (1995) acknowledges that a highly cohesive group (as was the case with NARC) is prone to making irrational decisions. Similarly, while they made policy decisions in the run-up to the elections, this group was insulated ‘from the outside environment’ and they did not consult the key stakeholders like the head teachers, teachers, education officials and even the parents – who would be the implementers of the FPE policy (Somerset, 2009). The group’s partial leadership was evident as they had vested interests in the final outcome of this policy – which would help them win massive votes due to the strategic importance of this policy. To this team, the policy was a means to an end. Finally, the homogeneity of the group members’ backgrounds, owing to the fact that they were all politicians from a wealthy background – most of whom had their children enrolled in international private schools both at home and abroad – the cultural disconnect meant that they did not share in the inherent needs of the poor people who were to be the beneficiaries of FPE. This lack of diversity in the group also meant that they couldn’t debate critical issues, as they had no contrary experience. These pre-existing antecedent conditions led the group to concurrence seeking, which may have produced symptoms of groupthink.
Most of the eight symptoms of groupthink, which show evidence of concurrence seeking, were also evident in this group. The NARC group’s popularity made them feel almost invincible, and their risky choice in a policy as hugely expensive as free primary education explains that. Without the required finances, the government had to rely on external donor agencies like the World Bank and DFID, which provided $50M and $21M respectively (Riddel 2003), to bridge the immediate funding gap. The group’s closed-mindedness is also evident when they fail to consult widely before deciding on the policy. Their main objective was to win, and they portrayed their opponents as terrible. They also felt pressured to stay in the group, especially as they had a ‘common enemy’ and also because they were working under stressful and high-pressure environment, knowing well that despite their popularity, the incumbent – a totalitarian President, could easily rig the polls in his favour. With all these symptoms present, there was a very high probability of groupthink occurring. Such pressures lead to irrational thinking and decision-making, which usually ends up in failure.
Janis (1982) lists consequences of groupthink, which are the symptoms of defective decision-making, which were evident in the FPE policy formulation stage. Defective decision-making here discourages the use of evidence to inform policy decisions.
These policymakers did not survey alternative policies or even objectives before settling on FPE. Abuya, Oketch and Musyoka, (2013) and Orodho (2014) found that the mass dropouts were due to poverty at the community and family level, which couldn’t be addressed by simply introducing FPE. Had the policymakers used evidence to inform their decision-making, they would have implement the programme in rural areas and urban slums where there is high poverty levels (Watkins, van Fleet and Greubel, 2012), rather than indiscriminately rolling out the programme countrywide.
Failure to examine risks of preferred choices was another symptom of defective policymaking identified in this policy. Although FPE was appropriate in addressing primary school enrolment in Kenya (UNESCO, 2005), free education was viewed by this group as the ‘magic bullet’ in achieving universal primary education (Abuya, Oketch and Musyoka, 2013). That however is not the case, and Kaimenyi (2016) rightly argues that school enrollment and actual learning are hugely different, yet the latter was not taking place. Supporting this argument is Uwezo’s (2015) annual learning assessment report found that learning levels are low, and that 30% of grade 3 pupils are unable to do basic grade 2 work and 8% of grade 8 (final primary grade) unable to solve the same.
Poor information search and selective bias in processing information (Jaris 1982) is evident, when the policymakers didn’t realise that Kenya has implemented a similar policy twice in the past – first in 1974 and in 1979. In both instances there was an initial massive enrolment, which was followed by equally massive dropouts (Somerset, 2009). This pattern is consistent with the FPE policy implemented in 2003. The policymakers did not seek to understand why the two previous policies failed in a typical groupthink fashion. The group also failed to work out contingency plans. Indeed, it can be argued that political narratives drove the FPE policy design (Uwezo, 2015) and its implementation was not well thought out and neither was there a sustainability strategy.
Indeed, there was a huge disconnect between the policymakers and the reality of the implementation, which ended up being the major shortcoming for this policy (Abuya et al., 2005). As noted above, the FPE policymaking process suffered from most of the symptoms of defective decision-making, which greatly increased its chances of failure (Janis, 1982; Hart, 1990). It can also be argued that the fact that implementers were excluded from the decision-making stage, suggests that the policymakers were not interested in the final implementation of the policy. It is worth noting that, although Radian and Sharkansky (1979) have argued for the consideration of implementation at the policy design stage, Ragan (1984) challenges this, arguing that considering implementation at the design stage compromises the policy goals to the interest of the bureaucratic implementers.
Despite its poor policy as a result of the defective decision-making, small working groups cannot be avoided, and due to their effectiveness, governments still greatly rely on them in policymaking (Parsons, 1995). By understanding the key sympoms of groupthink tendencies, analysts and policymakers are able to shield themselves from such, and in case they find themselves in groupthink settings, maybe remedy the situation.
Kenya’s Anti-Corruption Policy
The second policy this essay explores is the implementation failure of Kenya’s anti-corruption policy (Anti-Corruption and Economic Crimes Act) enacted in 2003.
Kenya has instituted several legal and institutional frameworks to fight corruption, These include the Public Office Ethics Act, 2003, the Public Procurement and Disposal Act, 2005, Leadership and Integrity Act, the Ethics and Anti-Corruption Commission Act, 2011 and the Public Procurement Oversight Authority Act among others. Beyond the State, several stakeholders are also engaged in the fight against corruption; including the media, civil society, private sector and donor agencies (Martini, 2012).
Yet, despite putting in place some of the most robust measures to curb corruption, endemic and systemic corruption still persists in Kenya and these strategies have borne no fruits (yet) (Hope, 2001). Indeed, Kenya is listed as one of the most corrupt countries in the world (Transparency International, 2015) and it is reported that an estimated one third of Kenya’s annual budget – about USD 6Billion (6B) – ends up in corruption (Miriri, 2016). The sheer scale of corruption scandals in Kenya is mind-boggling, and in 2016 alone, almost five major corruption scandals rocked the country. These ranged from the alleged misappropriation of USD 50M from the health ministry (Murumba, 2016); the loss of USD 18M at the National Youth Service (Guguyu, 2017) and USD 2.15B from Kenya’s controversial Eurobond funds floated 2 years ago, which cannot be accounted for (Wafula, 2016). The consistent pattern in all these cases is that no arrests or prosecutions have been made. In fact, Njoroge (2016) notes that less than 4% of the more than 1000 cases reported in government were actually recommended for prosecution. In a country where 34% of the population live below $1.90 a day (World Bank, 2015), these figures are not only astounding; but brings to question the government’s seriousness to tackle corruption, and raises concerns about the effectiveness of the anti-corruption policy.
A recent survey by Kenya’s Ethics and Anti-Corruption Commission found that 74% of Kenyans perceive corruption levels to be too high, and would like to see the perpetrators prosecuted (EACC, 2003). Anti-corruption efforts attract massive public support and one would expect the policy to be implemented with relative ease. Yet this is not the case. There are several reasons that can be advanced for this gap. These range from the lack of political will from the policymakers (who most are beneficiaries of corruption); the inherent ambiguity of corruption; as well as the intractability of corruption as an issue.
Corruption as a wicked problem is complex and ambiguous. When a policy concerns complex issues whose solution depends on the action and interactions of multiple stakeholders, there often tends to be high ambiguity and conflict (Matland, 1995). In the search for a proper theory to explain policy implementation process where there is a substantial conflict and ambiguity, neither top-down nor bottom-up is sufficient (Matland, 1995). I will therefore base my analysis of the anti-corruption policy’s implementation failure on Matland’s Ambiguity-Conflict theoretical framework.
The ambiguity-conflict theory gives a comprehensive approach explaining the gap between policy intention and implementation, by considering the extent to which a policy’s ambiguity and conflict impact its successful implementation. The core of the theory consists in four implementation perspectives: administrative implementation (with low conflict and low ambiguity) where resources determine outcomes, experimental implementation (with low conflict and high ambiguity) where contextual conditions determine implementation, political implementation (with high conflict and low ambiguity) where power determines the implementation and finally, symbolic implementation (with high conflict and high ambiguity) where strong coalitions shape the outcomes (Matland, 1995). In this case, the anti-corruption policy falls on the high conflict high ambiguity quadrant.
According to Matland (1995), ambiguity could point to unclear policy goals and means – or both, whereas policy conflict revolves around interdependence of policy actors, incompatibility of their objectives or a perceived win-lose element to their interaction.
Kenya’s anti-corruption policy has a visibly ambiguous goal of ‘ending corruption in Kenya’ (The Anti-Corruption and Economic Crimes Act 2003). Beyond this main goal, the policy also has other referential sub-goals. These are aligned to investigation, prosecution, prevention, education and asset recovery. The ambiguity in these goals leads to multiple interpretations, as different actors in the sub-system have different perceptions and definition of goals. This leaves the policy open to (mis)interpretation (Matland, 1995) and may affect its implementation. Furthermore, actors will form competing coalitions based on their chosen goal definitions. This is likely to lead to conflict, which can also hinder the successful implementation of the policy (Matland, 1995). He also notes that the strength of a coalition (whether in terms of power or resources) determines whether the policy would be successfully implemented. Yet, in Kenya, people who hold the power and resources may not have interest in reducing corruption as it benefits them (Hope, 2001).
But beyond the ambiguity of the goals, the definition of corruption in itself ambiguous, and may vary from one culture to another. For example, what is viewed as corruption from a western perspective may not be viewed as such in the Kenyan context. In a culture where ‘gift-giving’ is the norm normalised, there is a thin line between a gift and a bribe. Unfortunately, the policy cannot comprehensively define each type and act of corruption definitions, leaving this open to individual interpretation. This complexity makes it very difficult to fully implement a well-intentioned policy as the anti-corruption policy, leading to its failure.
The ambiguity in means allows corrupt powerful individuals to exploit policy gaps to evade justice. It also allows the process to be highly politicised, making the outcomes defined by the power relations between the policy actors in the sub-system. Despite the ambiguity in the policy goals and means, over time, these have become clearer, but the complete implementation is still lacking.
Lack of tangible implementation is common in this policy environment suggesting the need to reduce either the level of conflict or ambiguity in order to move the policy to a more successful implementation scenario. Ambiguous means therefore can be tightened, but goals should be left ambiguous, as corruption keeps morphing.
To ensure successful implementation in this policy scenario, Stachowiak (2016 et al.) recommend among others; capacity to engage opponents, consistent messages, transparency for winners and losers, tolerance for diversity, consistency over a long timeframe as well as knowledge capture and feedback
Conclusion
I have discussed two important policies in Kenya, whose implementation show a disconnect between policymakers’ intentions and policy implementation. The theories I have used are just among the many that can be used to explain policy implementation failure. In both the two cases, an clear emerging theme is the lack of policymakers’ political will and interest in the successful implementation of the policies –a scenario which makes failure inescapable.
It is important to note that it not always the case that all policies suffering from groupthink and ambiguity-conflict always end up in failure. But by understanding the key concepts, we are able to learn from past failures, and avoid the same mistakes in future policies.
The policies analysed above were top-down ‘heavy’, but it’s important to note that not all top-down policies are bound to fail, especially if implementation structures are legally structured so as to enhance the likelihood of compliance with policy implementation. Indeed, with more crisp policy instruments, sufficient resources and a reduced interface between policy as intent and the expected implementation, success rate can be increased.