Data Maturity

Menu
The Open Institute

Findings

Data Summary

No Data Found

Figure: Cumulative summary of all responses across all counties.
A large proportion of responses fall in the later stages of the model. However, a push is necessary to add more responses to the growing and later stages and out of the seedling stage of data maturity.

While the data quality and availability assessment will provide an indication of the robustness of data, it does not explain potential reasons for why this is the case and what should be done. The objective is to ascertain intentionality, ensure sustainability and scalability in strengthening the data ecosystem at the county government level for the particular theme.
According to the respondents, a trend can be seen that data quality generally mimics the institutional enablers that ensure that it continues to be produced and used. In other words, it can be seen that respondents felt there was least growth (or ‘growing’ stage) for both quality and causality enablers.
And even though there are a large number of respondents who cumulatively believe that quality data is both available and causality enablers present, that is, ‘flowering’ and ‘fruitful’; it is also seen that much still remains to be done as respondents also felt that data quality and causality enablers are at a ‘seedling’ stage.

Quality and availability

No Data Found

Figure: Overall Summary of responses on Data Quality and Availability.
The majority of responses fall within the flowering and fruitful stages, however the number of responses in the seedling stage is still high. This is especially true under data quality and availability where there are few and no responses in the growing stage, respectively.

The departments mostly scored within stages 3 (Flowering) and 4 (Fruitful) on data quality and availability, with a considerable amount of scores at the initial seedling stage. The scores in the data quality and availability section were analysed in four categories, namely data collection, reporting, availability, and quality. The results are summarised by category below and in more detail in the sectoral reports.

It is apparent that counties are collecting and reporting data on a regular basis for the sectors under study. However, the largest percentage of data collection is done manually, from primary data sources, by field officials and others charged with providing public services (for example health workers or agriculture extension officers) via various government institutions or at ward level and compiled into digital reports at the sub county and then county headquarters. We posit that some sectors such as governance and health where there have been strong national implementation structures and centralised national-level databases for data management have relatively standardised methods for consistent data collection and reporting. This includes the Government Human Resource Information System (GHRIS)4, National Education Management Information System (NEMIS)5, Integrated Financial Management Information System (IFMIS)6 and the Kenya Health Information System (KHIS)7. The result is that some departments register strongly on ‘flowering’ and ‘fruitful’ stages due to institutionalisation of processes, people and even resources for continued sustenance and improvement of data management practices and use.

Incidentally, during public meetings and public participation, data is collected using physical attendance registers with information captured including the demographics of the citizens that were present and these are stored in physical files and then summarised in digital formats for reporting.

Our interviews revealed that data is mostly reported in quarterly and annual reports for accountability purposes and most of these are available to the public. Most departments received high scores when it comes to reporting on data as can be seen in Figure 3.

The agriculture and education (focused on Early Childhood and Development Education (ECDE) and Technical and Vocational Education and Training (TVET)) departments within counties are devolved functions according to the 4th schedule of the Constitution of Kenya (2010) and so the responsibility of data management naturally also mostly falls on the shoulders of the county governments.

Data is collected on site through manual registers. They have a variety of systems for data storage in place which includes manual systems (that is physical files) to spreadsheets (on PCs and/or Google sheets). This differs depending on the priorities of the departmental managers, data competencies and the county government. Overall, it was found to be more difficult to standardise data collection practises within the agricultural sector because agricultural activities vary in different counties. Respondents revealed that most of the quantitative data is estimated based on formulas since the field staff cannot visit and measure every farm.

Most of the data has been made available to the public mostly in the form of PDF documents on county government websites and the websites for other national government agencies such as DHIS and KNBS, which is not a document format that citizens can easily extract data from for further analysis. In most cases, members of the public can request this data from local government officials. However, most of the data is available in county progress reports, including the occasional CIPD reports and sectoral reports.

The Office of the Auditor-General is mandated by article 229 of the constitution to audit and report on any entity using public resources for their operations. This includes county governments that are dependent on the national government for the majority of their resources8. However, according to the respondents, processes for third party validation of data are not universal and many departments do not have additional systems in place to complement the auditing process. The focus of data validation for most of the respondents was financial accountability rather than quality assurance.

Causality

No Data Found

Figure: Overall Summary of responses on Causality.
Scores are highest under incentives & accountability, and coordination & governance. There is room for improvement in the areas of non-governmental ecosystem and mandate. The widest range of responses was observed under resourcing, but with an increasing number of responses in the later stages of maturity.

Scores are highest under incentives & accountability, and coordination & governance. There is room for improvement in the areas of non-governmental ecosystem and mandate. The widest range of responses was observed under resourcing, but with an increasing number of responses in the later stages of maturity.

The study found that counties generally score highly on mandate, with most responses falling under the ‘fruitful’ stage of the model. This may be linked to the legal frameworks that are in place to ensure that counties collect data, including the statistics function that falls under the county planning and development role in the 4th schedule of the constitution of Kenya. The bill of rights also grants citizens the right to access information held by the state, and the state is required to publish and publicise important information in the same chapter.

The scores for incentives and accountability mostly fall within the last two stages of maturity. This may be due to the demand for accountability to the national treasury to receive funds. It can be attributed to the need by counties (and thereby departments) to comply with most of the reporting demands of the Office of Controller of Budget (OCOB). Disbursements to county governments are made by the National Treasury upon passage of an Appropriation Act, requisitions involving the Commission for Revenue Allocation (CARA), and upon approval by OCOB.

One common theme that stands out in causality ratings is the lack of resourcing – both staff and material resources – specific to data in county governments. The governance and health departments receive systemic data management support from the national government and this is reflected in their scores. However, much more can be done to streamline processes and investments in better data infrastructures to collect data in a harmonised format from all counties.
In the agriculture and education departments, each county in this study has developed their own data management system to meet the reporting requirements. The governance and health sectors collaborate with the national government for data management via the IFMIS system and DHIS, with standard protocols and formats for reporting clearly outlined. From the responses, systems for data privacy and security in line with the data protection act are not part of standard practice for most of the sectors under this study. However, these are considerations that are due to be addressed by the officials.
County governments have opportunities for collaboration with entities outside of local and national government, but these collaborations are often short-lived and nGOs do not cover all stages of the data management process. Some respondents also alluded to difficulties in collaborating with actors outside of government due to political interference.

GOVERNANCE
SECTOR

Often found under the Department of Finance and Economic Planning and/or under the Office of the Governor, a key function of the governance in county governments is facilitating of public participation. The Finance and Economic Planning Department also takes the lead on public finance, planning and monitoring and evaluation within the county government. The following indicators were included in assessing for data quality and availability for this sector, namely availability of:
  1. County budget – development and recurrent
  2. County annual reports
  3. Amount of ‘own’ revenue as a percentage of total budget
  4. Collection efficiency
  5. Total county employees per 1000 population, disaggregated by level of education (primary, secondary, university)
  6. No. of Services under E-Governance
  7. No. of town hall meetings held
  8. No of civic education events held in the County
  9. Annual number of public local government meetings and total attendance
  10. No. of policies and laws passed per year
  11. Does the county have an Ombudsman to address Citizen grievances
  12. Number of live broadcasts of House proceedings

Data Summary

No Data Found

Figure: Summary of Governance Sector Results.
The responses in the governance sector present an overall high level of maturity, with room for growth where both quality & availability and causality are concerned.

Results

The majority of scores when it comes to data quality and availability as well as causality under governance fall within the flowering and fruitful stages (chart on the right). Data is collected and reported regularly for most of the indicators under governance as part of county transparency and accountability processes as well as due to statutory reporting needs for independent constitutional offices such as OCOB, CARA, the Commission of Administrative Justice (CAJ), Council of Governors (CoG) as well as at the request of the national government.

Quality and availability

No Data Found

Figure 6: Data Quality and Availability Scores – Governance.

The scores are high across the board, with most responses falling in the latter stages of the model. However, there are more responses in the flowering stage than the final stage of maturity, and therefore room for further improvement.

The interviews and data show that budgeting and finance data is key for the governance sector, mostly for planning and accountability. All finance and procurement data must be logged in real time to IFMIS, which is also the e-procurement system used by both the national and country governments. Data is uploaded to the national database (IFMIS) in real time and hence they have efficient systems in place to manage collection and sharing of finance data. Details of procurement processes, payments and revenues are regularly logged on the digital platform. Additionally, most of the counties also possess Revenue Management Systems which are key in issuing business permits, tracking various payments made to the county by businesses. This includes mapping out where these actual businesses are and their contact details.

Traditional line-item budgets (including Kenya’s budget until 2013/14) previously focused on providing details on all the government was spending money on leading to voluminous data on specific inputs15. Post this period, there was a shift to program- based budgets (PBB) which organise the budget around objectives rather than inputs16. Data on/for public participation is key to drive this sector; PBB as driven through counties makes a distinction in earmarked development budgets to both flagship projects as well as an amount allocation to the ward level that must be spent on priorities identified by citizens through public participation. Data from public participation is essential for accountability processes and these exercises are well documented on paper at the time of the event.

The counties participating in this assessment have also instituted feedback mechanisms (similar to CAJ) and systems (for complaints, compliments) including suggestion boxes, online forms, access to local government officials at sub county headquarters, and town hall meetings. However, for most of the counties in this assessment, we found that the avenues for feedback are not anonymous as the views are collected at public events, offices or call centres and SMS systems which log personal contacts of reporters.

As stated, most data under governance is reported to statutory bodies, primarily OCOB (financial performance), CARA and CoG through quarterly and annual/quarterly reports are available to the public in PDF format in the websites of these bodies. Reporting to bodies such as CAJ, the National Gender and Equality Commission (NGEC) (see examples1718) is noted to be still lagging behind.
The data on public participation is mostly held in physical files using registers which are filled in manually in these meetings. The county assembly has its own data repository in the form of hansard reports and which are managed internally in each county. Most assemblies in the assessment counties do not publish hansard but there are committee reports available on county assembly websites that provide a lot of detail in PDF format. In summary, whereas Figure 6 above shows that most counties are receiving higher scores for data availability; there is a considerable portion of responses in the seedling stage.
Finance data is regularly audited by the national government. The results of these audits are published by the office of the auditor general19 and accessible to members of the public. As a result, we find that this data is of very high quality.

Causality

No Data Found

Figure: Causality scores – Governance.
This section has a high proportion of responses in the fruitful stage. There is a significant number of responses in the seedling stage, especially in the non-government ecosystem category.

Financial incentives and penalties are used to drive collection and reporting of this sectoral, but mostly budget related data. According to the Controller of Budget Act of 2016, the Controller of Budget shall, in accordance with Article 228 (6) of the Constitution, submit to Parliament quarterly budget implementation reports for the national and county governments within 30 days after the end of each quarter. The law further states that a public officer, State Organ or State office shall cooperate with the Controller of Budget to enable the Controller of Budget to carry out his or her functions. Those who refuse or fail to cooperate with the Controller of Budget as required by this section commits an offence and is liable, on conviction, to a term of imprisonment not exceeding two years or to a fine not exceeding one million shillings, or to both. At the time of writing, Members of Parliament have been pushing for harsher penalties for late fiscal reporting (a jail term of 5 years and/or fine of up to 10 million shillings).

County governments rely on legislation and infrastructure from the national government to support their data collection and right to information mandates. It should be noted however that the National Treasury has been faulted by Senators and the Council of Governors over perpetual late disbursement of cash to counties, a trend which the leaders say is slowing down the implementation of devolution.

All data on procurement, budget and revenue is recorded on the IFMIS platform in real time. Data collection outside of financial transactions and processes is mostly manual but transferred to digital platforms for reporting to the national government and in annual progress reports.

Data is a top priority for departmental managers under governance. This sector is heavily involved in processes related to revenues and budgets which points to a need for higher accountability via regular reporting of data. This department invests in data management with regular and structured budget allocations in place.
Governance departments employ dedicated staff for data management, including monitoring and evaluation and planning officials. These employees mainly deal with ensuring that data is collected from all departments and reported according to the schedule laid down either locally or nationally.
Most secondary data sources in this sector come from the national government via KNBS and other MDA’s such as the Communications Authority of Kenya (CAK).
Infrastructure improvements underpinned by availability of cheap smartphones and digitization of crucial government services such as e-Citizen services, National Transport & Safety Authority’s Transport Information Management System (TIMS), Kenya Revenue Authority’s iTax, Huduma centres, among others, have increased the appetite/demand for internet access due to availing of these services online in recent years. According to the Communications Authority of Kenya (Sector Statistics Report for Quarter 1 of Financial year 2021/22), by 30th September 2021, the number of mobile phone devices accessing mobile networks stood at 59.0 million, out of which 33.0 million were feature phones and 26.0 million smartphones (the penetration levels of feature phones and smartphones stood at 67.9% and 53.4% respectively). Still, according to the Kenya Economic Update by the World Bank (2019), whereas 44% of the urban population have access to the internet and only 17% of people in rural areas have access. In the counties of interest, it was found that people mostly relied on access to the internet through cyber cafes and at sub-county offices (as opposed to through their smartphones) when it comes to government e-services. Most people in the rural counties prefer basic handsets with batteries that last longer compared to smartphones with shorter battery lives; and since most simple phones do not connect to the Internet, the result is fewer people with Internet access. Additionally, gender power dynamics are another obstacle to Internet access in rural areas -with many rural women not participating at the household level. And remote areas in some counties still do not have the appropriate infrastructure for all residents to have access to internet services – which might hamper any e-government initiatives that are planned.

HEALTH
SECTOR

In line with the 4th Schedule of the Constitution of Kenya 2010, the health sector at the county government level is responsible for county health services, such as, facilities and pharmacies, promotion of primary health care, food safety and waste disposal, and other health and safety related responsibilities31. The following indicators were included in the quality and availability assessment for the health sector:
  1. Doctor:population ratio
  2. Health facility numbers in the County
  3. Proportion of deliveries conducted by skilled birth attendants
  4. Proportion of children under the age of 1 fully immunized
  5. Proportion of general population testing positive for malaria
  6. % of villages declared Open Defecation Free
  7. Number of environment & public health workers
  8. No of TB cases identified and put on treatment
  9. Proportion of pregnant women attending 4th ANC visit
  10. Proportion of women of reproductive age accessing family planning services
  11. HIV prevalence rate
  12. Percentage of children under 5 that are underweight

Data Summary

No Data Found

Figure: Summary of Health Sector Results.
Most of the responses from the health sector fall within the final two stages of the model.

Results

Most of the indicators on the questionnaire are included in the District Health Information System (DHIS) or Health Management Information Systems (HMIS) which are databases that the national government uses to collate health data from different health facilities across the country.

Data practises in health are boosted by the efforts of the national ministry and partners when it comes to health; to collect data across the country and make it available to authorised users on an accessible platform (KHIS). There is an existing universal system for monthly data collection and reporting with regular validation and auditing. These structures and investments reflect on the scores across the board for the health sector to mostly fall within the flowering and fruitful stages of the model for both quality/availability as well as causality. Moreover, please note that three counties (Nairobi, West Pokot and Mandera) results are not reflected in these results as it was not possible to collect data in these counties by the time of reporting.

Quality and availability

No Data Found

Figure: Data Quality and Availability Scores – Health.
The health sector scores highly in data collection and reporting. However, there is room for improvement in scores across the board with a large number of responses in the seedling and flowering stages.

Data is collected in (near) real time as part of health facility records at the individual facility level. Most of this data is collected in physical forms and transcribed into the information systems by health workers. Whereas this is occuring in public health facilities, county governments health departments also engage with private and faith-based institutions to request for similar data. Most of this might not be captured in HMIS but it is nevertheless used by the county governments when it comes to planning.
The database is updated up to the sub county level on a monthly basis with data from health facilities in all counties. All staff including location and specialisation and health facility data is stored on the HMIS database. The data also forms part of annual reports as these indicators are part of the CIDP agenda for most counties. These reports are available on most county websites in PDF format.
This data is presented on the KHIS portal which can be accessed by authorised personnel as needed and based on their roles.
Data each of the key indicators under study is recorded with location data down to the health facility level. Demographic data is included in health records and the data is audited often.

Causality

No Data Found

Figure: Causality scores – Health.
The health sector scored highly on incentives & accountability, and coordination & governance. Scores in the non-government system category are lower than for any other category under health.

Health facilities are required to report this data to HMIS regularly for compilation and reporting to the national government. However, from the interviews, it was interestingly apparent that most respondents were not aware of any clearly articulated penalties for not reporting the data beyond withholding resources.
Besides monitoring of the disease burden, the data is used for monitoring costs and revenues for health departments. Health departments therefore receive regular and structured budgetary allocations for data management and collection based on these needs.
Most of the data is collected manually at individual health facilities and reported regularly on the HMIS database. It was noted however that the health workers were also primarily responsible for also capturing data in the HMIS; as such may not have training in data-specific roles (including data analysis) as most of the staff carry out this work in tandem with their other responsibilities.
County health facilities are well connected to the national data network and they have clear guidance on collecting and reporting the data to the public and national government. Health departments work closely with non-governmental organisations, and regularly coordinate using their data in planning and operations.

AGRICULTURE
SECTOR

Similar to the Department of Health, and in line with Provisions of the 4th Schedule
of the Constitution of Kenya 2010, County Departments of Agriculture, Livestock
and Fisheries are responsible for crop and animal husbandry, livestock sale yards,
county abattoirs, plant and animal disease control, and fisheries32.
The following indicators were included in the quality and availability assessment for
the agriculture sector:
1. Tonnes of Maize Produced annually
2. Percentage of population receiving food aid
3. Proportion of animals vaccinated
4. No of annual trainings for farmers
5. Number of active farmer associations
6. Hectares of arable land under crop production
7. Tonnes of fish produced
8. No. of forests conserved, managed and protected
9. Ha of crops under irrigation
10. Number of beneficiaries accessing farm inputs (tools, seeds, fertilizers,
pesticides)

Data Summary

No Data Found

Figure: Summary of Agriculture Sector Results.
A large proportion of results fall within the final two stages of the model in both sections. Many responses under data quality & availability fall under the first stage of the model.

Results

Data for the indicators under the agricultural sector in this study is mostly collected manually by field officers and extension officers and thereafter calculated according to preset formulas (drawing from sampled data). Some directorates such as those dealing with food aid and forestry may fall under other departments (or sectors) and not Agriculture and so this study may have excluded some key indicators. In Agriculture, there is no single top-down Management Information System for data management from the national government so the systems vary according to the needs at hand. Some examples of such systems include: 1) The Digital Good Balance Sheet33, 2) Kilimo Open Data Platform34 among others such as the Livestock Identification and Traceability35 (LITS) which are being piloted by the Ministry of Agriculture, Livestock, Fisheries and Cooperatives36. Admittedly, some activities in this sector are undertaken more by private sector actors than the county government and often this data may not be collected. Examples might be Agrovets who also provide advisory services when they come to buy farm inputs. These and other factors lead to a higher proportion of scores in the seedling stage for this sector. However, the number of responses in the flowering and fruitful stages of the model remains high.   For the participating counties, data on fisheries is not available; on one hand fishing is highly privatised and the study could not find formal coordination mechanisms for data sharing in this sub-sector.

Quality and availability

No Data Found

Figure: Data Quality and Availability Scores – Agriculture.
A large number of scores in this section are in the seedling stage of the model. However, a good proportion of responses still fall under the flowering stage which signals growth.

The study also found that in most counties, for crop farming, data is collected by field officers who visit individual farms on a regular basis, whilst data on in person training and vaccination records are collected on the day of the event and stored in physical registers.
The systems in place for reporting agriculture data do not have static reporting requirements due to the nature of agriculture that may encompass different types of crops with different planting cycles; and different types of livestock which produce different products and serve different purposes. Agriculture is a mainstay of the economy in Kenya and data is key. Regular reporting occurs at the county and national levels; unfortunately, a lot of times the reports (annual/quarterly or other) do not include the exact figures.
The main priorities of the agriculture departments from the counties selected for this study cut across different indicators; they have different agricultural outputs and the indicators selected could not do justice to all the counties uniformly. But it has been indicated that the indicator set was too broad to adequately cover all the counties (and it is possible to extend this tool to capture more indicators). However, as has been indicated, agriculture is not just a mainstay but also a source of livelihoods in Kenya, contributing 26% to GDP directly and another 27% indirectly through linkages with sectors such as manufacturing.
In most of the counties under study, the majority of the figures on crop farming are estimates obtained from inserting this data into a formula. Often this is due to logistical constraints and lack of resources (including competent personnel) to regularly go to the field to collect this data. Also, for all the counties under study, most of the data in the agricultural department is not subject to independent process audits.

Causality

No Data Found

Figure: Causality scores – Agriculture.
The agriculture sector scores highly on mandate, incentives & accountability, and coordination & governance. Scores are more evenly distributed across the stages of maturity in the resourcing and non- government ecosystem categories.

The national government, more specifically, the technical working group/committee on Agriculture Statistics (in collaboration with KNBS38) requests data from the county level due to its mandate to undertake agricultural census, annual surveys and high frequency surveys. Data also flows up from the county level to the Ministry of Agriculture, Livestock, Fisheries and Cooperatives; but this does not occur in a systematic fashion because there is no harmonised Management Information System that handles the various data related to different agricultural value chains.
On one hand, data collected under agriculture is used to monitor costs and revenues. On the other hand, such data is also used to monitor productivity. The field staff that collects data forms a large part of the workforce; collecting the data is a specialised function needing understanding of processes and procedures.
In most counties, data in agriculture is collected physically at farms or in-person events and meetings. We found that in most counties, data storage is eventually digital – in staff laptops and county computers. However, this data is not centralised. There is also not enough breadth in terms of skilled staff complement to accommodate quality data collection.
This study found that in most counties, the agriculture departments are open to collaboration with non-governmental agencies for partnerships when it comes to specialised programs and studies.

EDUCATION
SECTOR

In Kenya, the government’s contextualization of education and training sector contains the following levels in terms of structure:

  1. Early Childhood Development and Pre-school Education
  2. Primary Education
  3. Secondary Education
  4. University Education
  5. Technical and Vocational Education and Training (TVET)
  6. Teacher Education and Training
  7. Non-formal Education and Adult Education
  8. Special Education

Similar to the Department of Health, and in line with Provisions of the 4th Schedule of the Constitution of Kenya 2010, the education department at the county level is responsible for Early Childhood and Development Education (ECDE) and Technical and Vocational Education and Training (TVET)40.

The following indicators were included in the quality and availability assessment for the education sector:

  1. ECDE enrolment rate
  2. ECDE teacher/Pupil ratio
  3. Rate of ECDE student transition to primary school
  4. Number of ECDE classrooms constructed and equipped in the last financial year
  5. Number of ECDE schools with a feeding programme
  6. Amount spent per ECDE student
  7. ECDE learners with special needs(%)
  8. TVET enrolment rate
  9. TVET instructor:trainee ratio
  10. TVET course completion rate
  11. Number of vocational training centres equipped with modern tools and equipment
  12. Proportion of TVET centres with access to internet connectivity

Data Summary

No Data Found

Figure: Summary of Education Sector Results.
There is a high proportion of responses in the flowering and fruitful stages of the model, but still a high number of results fall within the initial seedling stage.

Data management in the education sector is led by the national government through the National Education Management Information System (NEMIS). NEMIS is a tool that automates the efficient management of basic education in Kenya through collecting data and information from education institutions; processing and reporting the status of designed indicators; and providing the sector a solid ground for effective management to ensure that every learner counts. For example, the government has used the system for Form One admissions and in the provision of the Medical Insurance Cover for secondary school students. However, data collected at the institution level is mostly in physical records that are later uploaded to the database. Logistical hurdles such as connectivity issues and lack of trained staff are some of the barriers to schools accessing the system directly.

In this study, a larger proportion of responses fall in the first stage of the model, specifically in the causality section. However, due to the reliance on the national government for funding, the county education departments must collect and report data on students, instructors and resources regularly to avoid related penalties. Schools are also encouraged to record data in the NEMIS platform for each student for disbursement of per capita resources. This improves the scores for data quality and availability for the department.

Quality and availability

No Data Found

Figure: Data Quality and Availability Scores – Education.
The majority of cores are distributed between the initial seedling stage and the last two stages of the model. The scores are higher for data quality and reporting than other categories.

The government of Kenya is committed to providing education to all its citizens. When it comes to data; however, the number of school-age children who do not have access to education services remains high – especially among children with disabilities – more than those who do not have disabilities.

Education data is collected daily at the school level and reported to the sub county offices on a monthly basis – this data is vital in informing important decision making such as learner capitation grants43 as well as issues such as learner and teacher absenteeism. Additionally, in most of the counties under study, most children with special needs are enrolled in separate schools/classes.

Data on learners with special needs was not readily available. At the national level, we have the Kenya National Special Needs Education Survey – first unveiled in July 2016; but this has not been regularly conducted by the Ministry of Education.

For the devolved functions, data is collected at the individual school level and collated and reported at the sub county level. Whilst this data is stored in digital formats; it resides in departmental laptops/computers and there is no centralised database that houses this information.
At the local level, we learnt that feedback mostly occurs through school boards of management where parents, staff and management identify and solve issues. This means that there are opportunities therein to interrogate the quality of the data as it gets used so that it can be improved. Besides such feedback however, we learnt there are not many other avenues for citizens to provide feedback on the quality of the education services and such data is not intentionally sought by departments.

Causality

No Data Found

Figure: Causality scores – Education.
The majority of the scores in this section fall within the first and last stages of the model.

We learnt that county education departments are required to report this data for budgeting and other planning activities. However, the study revealed there is no standard or centralised county system/database for the devolved education services; individual county departments have created their own systems and templates in place for data reporting (that is, templates or approaches are also not uniform across board).
A large proportion of county staff in this department are involved in collecting, collating and reporting data from individual facilities through the sub counties to the county department. The staff are trained on data collection according to local manuals.
Data collection in the education sector is largely non-digital. Individual schools collect and store data in paper records which are manually compiled using templates by field officers/extension officers. Data storage is largely digital but mostly offline.
As stated, counties have developed their own templates for manual data collection that are used by field officials at schools, and at sub county level. We found that there are collaboration efforts by non-governmental organisations when it comes to supporting some services; for example, some organisations are involved in feeding and other programs in the education sector. Also, whereas the private sector is heavily involved in education, data from private institutions is not often included in the data collection and analysis.

Right to Information

The constitution of Kenya has an overarching legislation on the right of citizens to access information and this is largely acknowledged and upheld in county governments. In 2016 Kenya passed the Access to Information Act, 2016 in fulfilment of its obligations under the constitution Article 10, Article 33 and 35. The Commission on Administrative Justice (CAJ), the oversight mechanism under the Access to Information law, has also not filed annual reports to the National Assembly as required by law (during the period from 2016). Consequently, there is no report yet to enable an assessment on the state of access to information in Kenya44. In addition, Kenya is one of the 78 national signatories to the Open Government Partnership (OGP) and has been since 2011. At the time of writing, the country is implementing the Fourth National Action Plan with clear commitments around beneficial ownership and open contracting.

Data collection & storage

Data is largely collected manually and stored digitally in all departments, often in a decentralised manner offline on local laptops/computers. The exception is data related to finance and procurement data which is uploaded to the IFMIS database in real time as well as the Health data.

Feedback

Channels for feedback are not anonymous in most county governments, and data collection often stops at the submission. The feedback is directed to the department in question and followed up later in the process but this is often not recorded. Feedback is also welcomed during public meetings and recorded for further accountability. During the pandemic, counties have evolved to receive feedback through more diverse channels such as social media and website forms.

Prioritisation

Data quality and availability was found to be of higher priority for departmental managers than the political class. This has bolstered claims that there are tendencies to prioritise programs based on politics as opposed to data and evidence where there are competing interests.

On one hand, departmental managers are held accountable by local and national government on their progress and use of funds, hence they recognise the importance of collecting and reporting accurate data regularly. On the other hand, there seems to be a need for further sensitization and capacity building on how data can be shared and used by elected officials so that the case for data management can be made and data management and governance are prioritised during county budgeting processes. This is depicted in the figure below

Resourcing

Most departments do not have a dedicated funding stream for staff and resources linked to data management. In Figure 21 below, the agriculture and education departments have a higher proportion of their budget dedicated to data management as there are higher numbers of officers collecting data in physical forms and/or templates on the ground. Data is integral to the operations of the governance sector as budgeting and planning activities involve data analysis and most of these exercises are carried out by the officials in charge of planning and monitoring and evaluation under this sector.

More data staff in the governance and health departments receive training specific to data management according to Figure 22 below. This is likely due to the technical requirements of the national databases; sadly the expertise to mine and analyse the data is not thorough in most cases. This may also be linked to the minimal interaction with the database during data collection in the education and agriculture sectors, since the majority of the exercise involves manual input into templates.

Resourcing

Most departments either have policies in place to guide data privacy and security, or have intentions of putting guidelines in place for the same in the near future. Data privacy and security is especially important in the health sector with a large proportion of respondents stating that the guidelines are already in place.

Private Sector

The level of data maturity in the private sector at subnational level in digitisation of data management varies by sector (for the 4 sectors). The figure below demonstrates that data management in the private sector is more advanced in the health and education sectors, possibly due to the extent of privatisation of services in these sectors, the number of records kept and wider availability of resources for data management in individual facilities.

Data Science Expertise

The level of expertise when it comes to data science/data analysis is generally low. It is attributable to the focus that most departments have towards service delivery (and not really research and analysis). Sadly, such talent at the county level was found to be rare – with most who possess these skills opting to seek ‘greener pastures’ in better paying jobs in the private sector, CSOs and even academia. The study found that ata f within county planning departments are mostly graduates of economics, mathematics and computer science.