240 likes | 256 Views
This presentation discusses the Bank of Papua New Guinea's experience in collecting and utilizing economic and administrative data for policy formulation and GDP computation. It focuses on the Business Liaison Survey and the methods used to ensure data accuracy and availability.
E N D
Economic Administrative Data Availability and Extraction: The Experience of Bank of PNG ~ Regional Meeting of Heads of Planning and Statistics, Noumea, 12-16th July 2010
Out of Necessity • BPNG carries out surveys that are not conventionally within the domain of central banks. • Because statistics/data/information on the real sector are not readily or timely available.
Outline of Presentation • Business Liaison Survey • Employment Survey • Retail Price Index
Business Liaison Survey (BLS) • For the Central Bank to have an indication of what is happening in the real sector in terms of business activity. • Source of quantitative data on business activity at the industry and regional level. • Vital for GDP computation and an integral input into the formulation of fiscal and monetary policy.
BLS (cont) – Administrative Data, Method, Coverage • The administrative data- financial records of companies. • A survey questionnaire covering sales/income, opening and closing stocks, wages & salaries, other sources of income, costs, depreciation, profit, etc, is used. • BPNG uses mainly the sales/operating income data. • The survey covers about 500 companies in the private sector representing various industries (10 of the 12), consistent with PNG standard industry classification (PNGSIC 2000). • Initially, the completion of survey questionnaire by companies was done on voluntary basis, based on goodwill and trust between the Central Bank and companies. • Later on, the survey came under the Statistical Services Act (Chapter No.386), whereby all businesses receiving the questionnaire are required by law to provide the requested data by specified due date (enforcement is an issue). Survey is under both NSO and BPNG.
BLS (cont) : Questionnaire-each question is designed to capture an element of value-added output method of GDP • Q1a – operating income excluding GST • Q1b – investment income • Q1c - extraordinary income • Q1 di – price increase • Q1dii – volume increase • Q2a – opening stock • Q2a – closing stock • Q3a – wages & salaries • Q3b – interest payments • Q3c - depreciation • ..and so forth
BLS (cont) : Industries • Agriculture, Forestry & Fisheries • Mining & Quarrying • Manufacturing • Construction • Wholesale Trade • Retail Trade • Hotels & Restaurants • Transportation • Real Estate • Other Services
BLS (cont): Regions • National Capital District • Southern • Highlands • Morobe • Momase • Islands • Autonomous Region of Bougainville
BLS (cont) - Sample: Selection of Companies • The sample is selected using the company register, a registry from the Investment Promotion Authority, telephone directory (white pages). • All large companies (by employment level & revenue) are covered. • A random sample of small companies. • BPNG recognizes – could be a bias towards larger companies – has attempted to balance the cost of sampling smaller companies with the benefit of creating a representative sample. • Consideration for a representative sample for each region and industry- e.g., a company in a small town (Alotau) may not be a big company by national standard but is big enough for the small town and is making a difference in the locality so is included. • The sample is reviewed periodically- most recently in 2001 and 2006.
BLS (cont): Two Parts in Questionnaire • First part covers a realized quarter/year for actual performance. • Second part covers expected performance for a specified 6 months or 1 year period for BPNG internal forecasts purpose. • All monetary answers are in Kina terms.
BLS (cont): Process • Survey is carried out on a quarterly and yearly basis. • Survey Questionnaire is sent after a quarter/year ends. • To ensure that the data quality is good and reflect the actual performance of a company: - a guide that contains definitions of terms and explanation on what each question is about is used by both the Bank and the company; - certain check guides entailed in the questionnaire, e.g., if answers on income and costs are correct, the answer on profit should be consistent with income minus costs; - actual field visits are carried out, previously half yearly, currently yearly for all companies and six monthly for big companies. - where no visits are made, Bank staff query companies through phone calls (toll free lines are installed and used) when some answers do not make sense.
BLS (cont): Company Visit • Graduate staff of Economics and Research Depts visit the companies in the urban centres that are covered under the survey. • A two-way interview is conducted • Bank staff ask the company representative (usually an accountant) to explain their answers on why there has been an increase or decline in sales in line with general business trend of the economy or region specific factor, etc, or seek clarification if something is not clear or a figure is an outlier and does not make sense. • Company staff on the other hand ask the Bank staff about certain Bank or Govt policies that may be an hindrance to business progress, discuss constraining factors. • The visit and interview is conducted with or without the completion of the questionnaire. If the latter, the answers can be better appreciated or queried against what the company said when the questionnaire is eventually received. • Often the questionnaire is completed at the time of the interview so the visit is one way of securing a response and increasing the response rate. • The company visits enable Bank staff to get a better feel of business activity on the ground.
BLS (cont): Staff Level & Data Inputting • Staff in the Economic Analysis Unit of the Research Department are trained to record each company completed survey questionnaire when it is received. 2-3 graduate staff and statistical clerk are involved, with a senior research analyst full time on it, who reports to the Unit manager. • Each company’s answers are entered into the company’s slot in the database spreadsheet by industry and region. • If a company operates in more than one region and/or is involved in more than one industry, the inputted data would reflect this. • Again, even at this stage, Bank staff can query a company if there are questionable data that were not picked up earlier.
BLS (cont): Response Rate • Bank aims for 80% response rate or better before an analysis of the figures is carried out. • Casual staff are hired to follow up on non-responsive companies as a way of securing a good response rate.
BLS (cont): Analysis • From the inputted data, each industry summation of sales of all the companies in that industry and each region summation of all companies operating in that region are derived. • The growth rate in sales for each industry and region for a concerned quarter or year is then calculated when comparing with the previous quarter or year to see how industries and regions are performing. • An indication of real activity can be inferred when accounting for price changes to deduce whether the change is a price effect or volume effect or both. • To ensure that the comparison is about like terms, the comparing sample of companies in quarter/year two must be the same as that in quarter/year 1, meaning a company that responded in period 2 but not in period 1, or the other way around, is excluded.
BLS (cont): Outputs • Quarterly Economic Bulletin – Commentary on Business Activity (Real Sector) Section. • GDP estimate and projection. • An input into monetary policy and fiscal policy formulations. Treasury consults BPNG on the latter’s BLS findings. • NSO seeks clarifications from BPNG on sectoral/industry results for GDP compilation.
Employment Survey • Unlike in many economies where an unemployment rate is calculated, in PNG the Central calculates an employment index. • Because of the dual nature of the economy- urban formal and rural subsistence economy where people own land and are “self-employed”. It would be difficult to calculate and discuss unemployment. • So instead n employment index is constructed for the formal private sector.
Employment Survey (cont) • A questionnaire asking companies to simply report the total number of employees, including casuals, they employ in a quarter. • Same sample of companies under the BLS are covered in the ES. • Just like in the BLS, the answers are given by industry and region as each company is categorized into industry and region.
Employment Survey (cont) • Index on employment growth calculated by industry and region on quarterly basis. • As done for the BLS, the comparative total responded companies between two quarters must be the same. • Output: Discussion of employment growth trend in the Quarterly Economic Bulletin of the Bank. • Good proxy of real activity growth. • Consistency check between employment and sales. • But sales reporting in QEB is one quarter behind employment.
Retail Price Index (RPI) • Monthly survey major supermarkets (shops) in Port Moresby. • The supermarkets are asked to state their monthly prices of the common grocery items that are included in the CPI basket. • The retail price figures are entered on a form through the means of bar codes scanning. • Also, the Research Dept receives from (I) NSO, the weekly market prices of the main fruits and vegetables that are included in the CPI basket and (ii) from ICCC the monthly fuel prices. • The data from the supermarkets are entered for each consumer item by supermarket in the database spreadsheet.
RPI (cont) • Staff follow up supermarkets that do not respond on time, and query volatile numbers. The supermarkets are contacted to explain figures. • In most cases verifications are given by the supermarkets and good economic reasons are given. • Price growth rates per item are calculated and aggregated into one worksheet. • Price changes can be worked out for each item, groups of like items, and for the overall index between months.
RPI (cont) • The purpose: for the Bank to have a more timely and regular information and indication of price changes of consumer items to aid its monthly monetary policy decision as CPI is released on a quarterly basis. • Although the information is for Port Moresby only, it is found that there is a strong correlation between the RPI and CPI (which is for the major urban centres). • So the Bank is comfortable in using the Port Moresby monthly RPI changes as a good guide to consumer prices changes for monetary policy decision. • BPNG does not publish it, for internal use only. Does not replace CPI. It is careful on this because its mandated monetary policy objective is price stability and conflict of interest must be avoided.
NSO is the official source of GDP and CPI. • Treasury Dept is the official source for GDP projection as contained in the Annual National Budget Documents. • Both institutions would have in some way have embedded in their figures some results of the BLS. • In all, there is a three way dialogue between NSO, Treasury and BPNG helping each other out.