Philippines Discussion Notes

Yasuhiko Matsuda (EASPR)

Victor Dumas (PRMPS)

The World Bank

1. 
   Good statistics — meaning accurate, timely and useful statistics — are important for achieving the development targets in the Philippine Development Plan (PDP). Without good statistics, targets such as “accelerating growth to 8 percent by 2010 and reducing poverty incidence among families by 10 percentage points” would be difficult to achieve, in part due to the inability of the government to make appropriate policies and allocate scarce resources to the sectors yielding the highest returns or offering the best prospects for reducing poverty.⁰ For example, weaknesses in the Philippine System of National Accounts (PSNA) have reduced the ability of the government to determine priority sectors to support. Moreover, the absence of reliable provincial-level poverty statistics and the late release of national and regional poverty statistics have limited the government’s ability to design better poverty reduction programs.⁰
   

2. 
   Good statistics are crucial in supporting the government’s social protection program. With a limited budget and large number of poor and vulnerable people, accurate, timely, and disaggregated statistics are needed to ensure that limited resources are well-targeted to the people who are most in need. In the past, the lack of accurate provincial and municipal-level poverty statistics and a good proxy means test have led to large leakages in the government’s various social protection programs.⁰ This has since been rectified by the development of a good proxy means test using municipal-level poverty statistics. The government’s conditional cash transfer (CCT) program offers an example of how good statistics can improve resource allocation and help bring down poverty rates. In recent times, the increasing frequency and magnitude of calamities and disasters have made the production of good statistics even more urgent. When typhoons Ondoy and Pepeng hit the country in 2009, the lack of accurate spatial data slowed down rescue and recovery efforts in the most affected areas. Ensuring the availability of good statistics is essential to improve disaster preparedness, crisis monitoring, and the design of rapid response mechanisms.
   

3. 
   Finally, good statistics are needed for good governance. A government that seeks to be transparent and accountable to the people must produce good statistics to back its claims. Bad statistics can mask inefficiencies in public expenditure and increase opportunities for corruption.
   

4. 
   The Philippine Statistical System (PSS) is the government-wide system for providing statistical information and services. It has a decentralized structure composed of six major statistical agencies (MSAs) and other units in the national and local government engaged in the production of statistics.⁰ A total of 19 government agencies support the delivery of about 60 designated statistics which include, among others, the estimation of the national income accounts and poverty, regular monitoring of prices, and various economic and social censuses and surveys.
   

5. 
   5. The PSS is regarded positively by the international community. The PSS ranks high in the 2009 Statistical Capacity Indicators of the World Bank.⁰ The IMF’s 2004 review of the Philippines’ observance of standards and codes found that Philippine macroeconomic data exceeds international standards of good practice in terms of the coverage, periodicity, and timeliness.⁰ Dissemination practices are consistent with Special Data Dissemination System (SDDS) requirements and the PSS as a whole exhibits a high degree of professionalism and integrity. The competency of Philippine statisticians is also well regarded in the region. Several countries have sent delegations to learn from the PSS and have benefited from the advice of leading Philippine statisticians on how to improve their statistical systems.
   

6. 
   There are several areas for improvement, however, starting with the need to improve inter-agency cooperation and focus more on developing good source data to enhance accuracy and reliability. According to the Report of the Special Committee to Review the PSS (2008), the decentralized nature of the PSS has spawned problems of overlapping functions, duplication of efforts, and inconsistencies in government-produced statistics (). These problems arose because departments do not use harmonized data collection procedures and standard definitions. At the local level, specific data collection activities are undertaken according to the priorities and needs of the local governments for planning purposes, but not necessarily in coordination with national agencies or following prescribed standards and guidelines.
   

7. 
   Longer-term statistical planning is not given enough attention and priority. The NSCB prepares the Philippine Statistical Development Program (PSDP) at the start of a new administration every six years to outline the directions, thrusts, and priorities of the PSS and to help the government achieve its targets. PSDP preparation and implementation is generally underfunded in the Budget, and high level officials are not actively involved in this process.⁰ Furthermore, the PSDP does not benefit from active participation of all the statistical agencies and government departments. Attendance in the NSCB Board is weak (Report of the Special Committee to Review the PSS 2008) and some agencies say that they are not consulted enough.
   

8. 
   Although the Philippine System of National Accounts (PSNA) is regarded as the most important measure of the country’s economic performance, there are questions about its integrity. A recent assessment (Harvey 2009) revealed that the integrity of the national accounts is compromised by three major issues. First, there are discontinuities in the national accounts series which renders the GDP series non-comparable over time. Second, the PSNA uses outdated benchmarks and parameters that do not reflect the current state of the economy. For example, it still uses 1985 as base year (unlike most other countries that use a more recent base year or have transitioned to chain volume measures), an industry classification as of 1977, and sector parameters from the 1980 Census of Agriculture and 1988 Census of Establishments. Third, data from the national income accounts do not correlate well with other official statistics, such as the foreign trade statistics and the balance of payments. This reliance on weaker data raises the risk of making wrong decisions that can harm the economy.⁰
   

9. 
   Many surveys that produce key economic and social indicators have problems of coverage, timeliness, and accuracy. Some statistics, such as poverty statistics and estimates of income and expenditure, are inadequate for their intended purposes because they (i) do not meet the desired level of disaggregation needed for planning, (ii) are released with significant lags, and (iii) do not capture some of the key social indicators that are essential to poverty monitoring. Data users complain that these statistics are often released too late to be useful for timely and appropriate inputs to policy making. Significant lags in the delivery of pertinent statistics have constrained the government’s responsiveness to address pressing challenges such as the recent crises and calamities.
   

10. 
    Expenditure compression in the national government has led to a significant decline in the real budget for the generation of statistics. On non-census years, the total budget allocated to the PSS averaged 0.5 percent of the national budget and this has declined further since 2001 (NSCB 2005a). The budget for statistical operations has been cut or delayed in many occasions, resulting in the delay or cancelation of crucial surveys and censuses.⁰ The budget for research, training, and product improvement was also totally removed, making it very difficult for statistical agencies to improve their capacities and products amid the growing demand for statistics, especially sub-national statistics. This resulted in a statistical system that is very hard-pressed to deliver its mandate given poor quantity and quality of resources.
    

11. The government’s tight fiscal policies also have adversely affected the number and quality of staff in the statistical agencies. Regular staff positions in the PSS declined by about 25 percent in spite of higher demand for statistics. The rationalization plans of some statistical agencies have not been approved, which has delayed the filling of key positions. In other agencies, the approved rationalization plans contain more cuts in positions than what the agencies had originally proposed. In addition, the technical expertise of personnel has declined because of the lack of training and skills development, not to mention turnover. An underlying reason for the low staff count is low compensation and weak career opportunities for statisticians, especially for senior level statisticians. On average, government compensation for experienced statisticians is about a third of what is offered in the private sector and by development agencies. This has led to frequent turnover and loss of seniority in the PSS.

12. 
    Recognizing these problems, the government commissioned a special review of the PSS in 2007 — the first major review in 22 years. A committee of experts was tasked to evaluate major issues and concerns in the current structure of the PSS and the functions and mandates of the MSAs vis-à-vis the quality and integrity of their products and services, and provide recommendations on how the system can be improved. The findings of the committee highlight the growing difficulties in the current statistical system. Among others, the review recommends the reorganization of the PSS into a single agency to “improve its efficiency and effectiveness in meeting demands for statistics.” In addition, it recommends that the budget for designated statistics be automatically appropriated and the research and training arm of the PSS be strengthened to address the increasing need for improved statistical products.
    

12. 
    The major statistical agencies have begun to work towards improving key designated statistics. In 2008, NSCB and NSO embarked on a project to improve the quality and usefulness of the National Accounts, in particular adopting the recommendations of SNA 1993 and 2008 and moving towards chain volume measures. The quality and coverage of surveys are also being improved by NSO, beginning with a comprehensive review of the household surveys and the development of an enterprise survey. To address the demand for more disaggregated statistics, city and municipal level poverty estimates were produced for 2000 and 2003 using the latest available census and surveys. Since their release, many LGUs and national government agencies have used them to improve the formulation of local development plans, improve targeting of social assistance programs, and in monitoring progress towards achieving the MDGs at the local level (NSCB 2009). NSCB has also begun preparatory work to formulate the Philippine Statistical Development Program (PSDP) for 2011-2017.
    

12. 
    To overcome the challenges, a coherent set of policy actions is needed. Foremost, coordination among statistical agencies needs to be strengthened, preferably via a unifying legal framework. This should be followed by formulating a realistic Philippine Statistical Development Program (PSDP). These two foundational reforms would then help the statistical system improve its resource pool. With more resources, the statistical agencies can then move to improve the quality of statistics.
    

12. 
    To improve coordination in the PSS, a first best solution would be to enact a unified legal framework that consolidates all existing laws governing the production of statistics.⁰ Having a unified framework for statistical operations can significantly improve horizontal coordination and institutional linkages among statistical agencies as well as vertical coordination within statistical agencies and the government bureaucracy. Such a law could also specify clearly how various government agencies could cooperate especially on data sharing. For instance, the law should mandate all government agencies to share data with NSO and NSCB, subject to provisions of existing laws, for national accounting purposes.⁰
    

12. 
    If a new legal framework cannot be realized in the short-term, then a second best solution would be to enact an executive order which would mandate free sharing of data across all government agencies. This will avoid instances when NSO, for example, has to pay a fee to get data from another national government agency, such as the Securities and Exchange Commission (SEC) or when NSO and NSCB have to enter into a memorandum of agreement (MOA) with other agencies to be able to acquire their data. This will not only promote greater appreciation and utilization of officially produced data but also build internal demand for statistical products in the formulation of government programs.
    

12. 
    Based on the findings of the PSS review, a bill is currently in congress that calls for a centralized statistical system. This proposal needs to be reviewed further. While the proposal appears to be well founded on the need to improve coordination among statistical agencies, such a move does not guarantee better results in the PSS. International experience on this issue is mixed and provides no clear-cut evidence that a centralized statistical system works better than a decentralized one. Countries with a more centralized system, such as Australia, Canada, and the Netherlands, perform as well as countries with a decentralized statistical system, such as the US, UK, Japan, and Indonesia. In many countries, good performance is conditional on effective implementation of current statutes and the availability of adequate resources and not solely on how the statistical system is organized. The proposal would benefit from a more in-depth study, with the aid of international advice, on how the PSS can be better organized.
    

12. 
    High level support needs to be given to the formulation of the Philippine Statistical Development Program (PSDP) and the active participation of key stakeholders, both data users and producers, in the PSDP process ensured. Given the resource constraints amid the growing demand for statistics, strategic planning in the area of statistics is essential in ensuring that scarce resources for statistics can be allocated efficiently and decisions on which statistics to produce, what methods to use, and how data are to be disseminated, can be made optimally. The experience of other countries, such as Indonesia, shows that well prepared plans raise the profile of statistics, increase the likelihood of getting adequate resources, and often translate to better performance. To build up the profile of statistics in the Philippines, it is important that high level officials, in particular key members of the NSCB Board and the NEDA Board, actively support the formulation of the next PSDP. Support can come in the form of active participation during consultation forums and in the form of budgetary support to improve both the development phase (2010-2011) and the monitoring and evaluation phase (2011 to 2017).
    

12. 
    The list of designated statistics needs to be reviewed and updated with the identification of key ones that should be protected at all times. The current list consists of about 60 designated statistics but there are some inadequacies. For example, the list does not provide statistics on Small Area Estimates (SAE) of poverty and data needed for rapid disaster or calamity response, both of which are very important to support the government’s social protection program. Meanwhile, new and important surveys, such as the enterprise survey and the survey of the informal sector, are not on the list. Furthermore, as discussed above, even within the current set of designated statistics some important surveys and censuses were not conducted due to budget shortfalls. This undermines to varying extents the ability of the government to make good policies and quickly respond to emergency needs, such as disaster and crisis management. The food crisis in 2008 is a case in point.
    

12. 
    Protection of key designated statistics could come in the form of “automatic appropriation”; that is, by exempting the survey/census budgets from general budget cutbacks, should they occur. If the budget for key designated statistics is not adequate to meet minimum quality standards, then it needs to be increased. To ensure that an adequate budget will always be available for the key designated statistics, the PSS needs to continue enhancing its medium-term expenditure budgeting. Table 2 proposes 16 statistics, surveys, and censuses that can be considered key designated statistics.