Linking Administrative Reform to Economic Growth patterns: review of methods and findings
Discussion Paper
Poverty Reduction and Economic Management Network
World Bank
February, 2005
Context
The Government of the Russian Federation has embarked on an ambitious set of reform programs that together are expected to overhaul the structure and operation of the state administration, and to enhance its performance. The Civil Service Reform program (2003-2005), the Concept of Reforming the Budget Process (2004-2006) and the impending Administrative Reform Program have been designed to achieve this objective. This brief note aims to highlight some potential tools that may be available to track progress in the reform process, as well as the impact of administrative reform on economic development.
The complexity of identifying the impact of administrative reform
There is a broad set of literature that explores the correlation between the quality of governance and economic growth, which with few exceptions establishes a strong positive correlation between the quality of Governance and levels of economic growth.[1] It is difficult to express the strength of this correlation in concrete terms, as the measurement tools are based on complex combinations of sets of indicators (e.g. one standard deviation on the regression leads to x percent of additional growth). However, even though there is ample evidence in the literature of a positive correlation between the quality of governance and economic growth, this evidence is mainly based on an analysis of large cross-country samples over a long period of time. Adapting such approaches to a single country in a medium term context is therefore a complex undertaking.
Constructing a model for analysis and tracking purposes is further complicated by ongoing discussions on the appropriateness and comparability of governance and administrative quality indicators. General macro-economic indicators that at first glance appear relevant and lend themselves to comparison, are easily distorted by national specificities, or are not necessarily good indicators for qualitative change. More complex measurement requires econometric analytical techniques and is difficult to use in a single-country context. Finally, whereas deeper qualitative assessment work on governance and administrative reform is often available, developing causal relations between such assessments and economic development in individual countries is an option that is rarely explored. Any attempt to build a functioning evaluation and tracking system for a specific case would therefore need to be based on a fair bit of ground work.
Structure of the Paper
This brief discussion paper starts with a review different types of indicators. It focuses both on the difficulty of identifying approaches to the measurement of the quality of governance, broadly, and administrative reform, narrowly, as such, as well on approaches to exploring the linkage between the quality of governance and public administration and economic growth.
In order to do this, we will first review different approaches to measuring progress in the development of governance systems and administrative reform. In this regard, three types of approaches are considered:
· Evolution of macro-economic and fiscal indicators, in particular the size of the wage bill and public sector employment levels
· Quantitative (econometric) analysis building on datasets of governance indicators, both externally and internally provided
· Qualitative assessment of administrative capacity and performance, baselines and Common Assessment Framework
This is followed by a discussion on the linkage between the quality of governance and administration debate to economic growth and development patters. In this regard, two approaches will be discussed, (a) econometric analysis based on long term observations in quality of governance and economic development patterns and (b) analyses that more specifically explore the link between quality and predictability of public administration and economic development patters.
Building a model that would specifically explore and track the linkage between improved governance/public administration and economic development in Russia would need to build on a combination of all or most of these approaches, and would require a more thorough discussion on how each of them can be made relevant to a single country context in a medium term time perspective. This mainly because most of the models and approaches discussed here are generally used for longer term and broad comparative work.
In view of the above, building a tracking and measurement approach of the type considered would be possible, but would require significant additional work, involving both Russian experts on administration and governance issues and international experts experienced in using measurement, assessment and modeling approaches.
Indicators of progress in administrative reform and quality of governance:
Fiscal and macro-economic data
Straightforward macro-economic indicators, such as the size of the public sector wage bill, and wage cost as a proportion of public expenditure, have often been used as initial benchmarks to assess the size of the administration and the impact of reform programs over time. Alternatively, comparative assessments of public sector employment are also used, though these are less usable as finding common definitions for categories of public sector employment across states is highly time-consuming and in many instances does not lead to relevant conclusions.
Wage bill indicators
Apart from the problems with the interpretation of data and differences in structure of public expenditure in individual states, one of the key problems in using wage bill dynamics is that they are based on the assumption that a reduction in wage bill expenditures is automatically a positive impact of a reform program. In this regard, states with the lowest ratios would stand out as success cases. In reality, the correlation between low wage expenditures to GDP and levels of economic development is not as straightforward.
Frugality on wage expenditures can easily lead to erosion of the quality of the civil service, especially in countries with a high growth rate, as the public sector can rapidly become uncompetitive compared to the private sector. In this case an increase in wage expenditures, within limits, would be the right administrative reform strategy.
Recent discussions in Kazakhstan, for instance, which has adopted a policy of moving towards a Singapore style civil service wage system over time, provide an illustration of this dilemma. Whereas Kazakhstan has a low wage/GDP (3.8%) and a low wage/total public expenditure ratio (16.6%), the civil service and public sector at large are suffering from high turnover and a rapid loss in qualified staff.
On the other hand, an over-expanded civil service and public sector wage bill is likely to constitute a brake on economic development, which is often illustrated by cases such as Mexico, Venezuela and other oil-rich states, where expansionist wage policies on the back of increasing oil revenues created a severe ‘Dutch Disease’ type impact in the 1990s.[2]
A further problem with the use of wage bill indicators is that it does not account for country specific features, such as size of the country, population density etc., in terms of the need to deliver public services to all. Therefore, looking at absolute numbers on the proportion of the wage bill and using them for comparison without factoring in such features may also lead to erroneous conclusions.
Finally, difficulties of interpretation remain when judging the size of the public sector wage bill in relation to GDP. IMF data (www.imf.org), which together with the World Bank’s database on public employment (http://www1.worldbank.org/publicsector/civilservice/data.htm) remain the main sources of information, do not distinguish between direct and indirect expenditures. For instance, the Southeast European Public Expenditure Review (World Bank, 2005, to be published) shows that including or excluding hospital staff and other health workers makes a difference of up to 3% in the size of the wage bill as a proportion of GDP. Comparator data therefore need to be used with extreme caution.
Using wage/public expenditure data suffers to some degree from the same problem as wage/GDP ratios. Comparability is limited due to differences in interpretation of what type of wage expenditure is included.[3] However, in the case of this indicator there is one absolute element of measurement: a high wage bill/public expenditure ratio crowds out public investment and thus is likely to have a direct negative impact on economic development, either directly, through reduced demand levels, or indirectly, by reduced investment in sectors such as health and education.
The wage/expenditure ratio would therefore be a more reliable indicator to judge the macro-economic impact of administrative reforms, though even in this case it remains difficult to establish clear boundaries between what is ‘reasonable’ and ‘excessive’. The latter is determined in part by the revenue/GDP ratio and in part by the stock of infrastructure and facilities in a given state. For advanced middle income states, for instance, ratios range from 12.4% in South-Korea to over 28% in Chile.
Based on the above considerations, the conclusion would be that whereas wage bill indicators can be used as a way to identify excessive spending patters, on both types of measurement discussed here, using these indicators to track progress in administrative reform in the Russian Federation would not be advisable. Analyses carried out by the World Bank[4] would lead to the conclusion that what is needed is an investment in wages, which is likely to drive up rather than reduce the wage bill. Whereas it is important that this is done in a sustainable manner, using a wage bill level indicator to track administrative reform would not be productive under these conditions.
Finally, the above reservations regarding the use of wage bill data also reduce the value of this indicator as determining factor in economic growth. States that severely under-invest in public sector wages may have a good wage bill/GDP ratio, but this is unlikely to translate in higher levels of economic growth, as underpaying public sector workers not only reduces the quality of the civil service and increases the risk of corruption, but also erodes the quality of health and education, which are both undisputed long-term determinants of economic growth.
Public employment levels
Even though there are some serious issues with the comparability of wage bill data and the interpretation in wage bill trends, as discussed above, wage bill proportions of GDP and expenditures are still generally used for comparative purposes. If such data are used as one part of an assessment, it would be advisable use public employment levels to control the distortions inherent in just using wage bill data.
A combined analysis of wage bill and staffing levels could help making a case for progress in administrative reform; the logic here is that higher wage bill for fewer employees would be likely to generate a competitive civil service, while a higher wage bill for the same, or more employees would be less likely to enhance competitiveness and quality.
However, assessing employment levels in a comparative perspective poses additional problems of reliability of data. The two main data sources for public sector employment data are the OECD and the World Bank database mentioned above. The ILO does not provide specific data on public sector employment. OECD data (OECD states only) are not available beyond 2000 (see http://www.oecd.org/document/1/0,2340,en_2649_34139_2408769_1_1_1_1,00.html), and are sketchy at best, most OECD states do not report, and employment numbers refer to overall employment in the public sector only, without making the necessary breakdown in categories of public sector employment. The World Bank’s database does provide a breakdown between civilian and non-civilian employment, as well as a breakdown in sectors inside civilian employment. However, data are also here incomplete and have not been updated since 2000, severely limiting comparability.
In conclusion, whereas in principle a combination of monitoring employment indicators in the core administration combined with monitoring wage bill development would be a good tool to come to a more nuanced judgment on macro-level impact of reforms than using wage bill indicators alone, a problem remains with finding benchmarking material. The World Bank database on public sector employment in principle provides the appropriate breakdown needed to conduct a sector by sector analysis, but the incomplete nature of the available data, and the time lag[5] pose problems in terms of the robustness of any analysis based on these data.
Governance indicators
Governance indicators have been a subject of discussion for the last decade and extensive work on these has been done, in particular in the World Bank. Two key reference documents are ‘Assessing Governance’ (http://www.worldbank.org/wbi/governance) and ‘Second Generation Governance indicators’ by Knack, Kulger and Manning (International Review of Administrative Sciences, Vol. 69, 2003).
The now generally accepted set of governance indicators defined by Daniel Kaufmann measures six dimensions of governance;
- Voice and Accountability
- Political Stability and Lack of Violence
- Quality of the Regulatory Framework
- Government Effectiveness
- Rule of Law
- Control of Corruption
Each of the six sets of indicators is scored based on measurement based on a large number of secondary sources (see http://www.worldbank.org/wbi/governance/pdf/govmatters3.pdf, pages 4-8) as well as targeted surveys and measurements of perception conducted by risk assessment firms, EUI, universities and polling agencies such as Gallup etc..
A comprehensive measurement on each dimension may not be required for the specific task of assessing the impact of an administrative reform program, as the key dimensions in this regard would be mainly Government Effectiveness, Control of Corruption and Quality of the Regulatory Framework (measuring the quality of the policy outcomes produced by the administration). Since the assessment model allows for both separate assessment and comprehensive assessment, it is possible to track only selected dimensions of governance, depending on the needs and purpose of the assessment.
The main drawback of this specific mode of governance assessment is that it is constructed to measure countries’ governance performance in relation to each other, rather than help tracking progress in a given country over time, the tracking system measures countries’ relative position over time.
The second main problem with the method applied is that it draws almost exclusively on subjective and perception based indicators, which enhances the risk of bias in ratings (the most common one is that countries with a strong level of economic development often are perceived as having higher quality governance systems, which is reflected in a bias in ratings), and would in addition make it easily susceptible to criticism inside individual countries.