Improvement the methodology and quality of measurement of the Producer Price Index (PPI) in industry from the perspective of data users- some reflections

Abstract

The development of ICT technology and digital economy contribute to changes in the way of production and redistribution of goods and services in various areas of life, which affect the growing expectations of users in relation to producer price statistics in the context of Producer Price Index (PPI). It is important to provide data on producer prices that reflect current price trends in a timely and accurate manner. They should be harmonised, comparable and ensure consistency of statistical observations with other coherent data. An important value of statistical data is therefore the optimal use of their potential in the processes of monitoring producer prices and assessment of their impact on the development of inflationary pressure on the producer’s market as well as control of the business cycle. What does the level of credibility of the PPI determine? How users can contribute for obtaining relevant data with good quality? What role do the methodology and quality of the survey process play in this respect? The answer to these questions is the article “Improvement the methodology and quality of measurement of the Producer Price Index (PPI) in industry from the perspective of data users- some reflections”.

Keywords

PPI user needs quality dimensions validation process

1. Introduction

The development of information and communication technologies ICT and the digital economy contributes to changes in the way of production and redistribution of goods in various areas of life, which affect the users’ growing expectations towards producer price statistics. Providing producer price data reflecting current and timely price trends is becoming increasingly important. Data should be harmonized, comparable and ensure that observations are consistent with other coherent data. Therefore, the optimal use of their potential in the processes of monitoring PPI data, assessing their impact on inflation pressure on the producer market and in control of the business cycle, also in the context of the occurrence of negative phenomena (e.g. related to a pandemic, recession), becomes a significant added value.

The needs of users and their impact on the model of price statistics always play an important role in this respect. They could constitute a source of inspiration for creation of added value in producer prices statistics. Lessons learnt from users expectations build a fresh outlook and way of thinking on improvements in statistics of PPI to better reflect specificities of innovation economy and development of Industry 4.0.

The aim of this paper is to present factors that could contribute to improve the methodology and quality of measurement of the Producer Price Index (PPI) in industry. In particular, there are discussed: PPI in industry from the perspective of data users, the methodology followed by Statistics Poland to compute PPI in industry, improvement of the quality measurement of the PPI in industry, methodological challenges and dilemmas in the face of the development of industry 4.0. Looking at PPI in industry from the impact of different factors may become an inspiration for new methodological works on improvements in the area of this statistics.

This article formulates only general questions and problems and indicates also general directions to find solutions in order to develop and to maintain discussion among statisticans on different, international forums how to adjust construction of PPI in industry to contemporary challenges and expectations of users.

The opinions presented in this paper express only the author’s point of view.

2. PPI in industry from the perspective of data users

The producer price index in industry is one of the basic economic indicators of PEEIs and indicators of short-term statistics STS1 and statistics within the Framework Regulation Integrating Business Statistics (FRIBS). It reflects actual price changes, i.e. changes in the structure of assortment as well as changes in the sales structure on the domestic and non-domestic market by sections of Statistical classification of economic activities in the European Community NACE rev.2 (equivalent to the Polish Classification of Activities PKD07): B-Mining and Quarrying, C-Manufacturing, D-Electricity, gas, steam and air conditioning supply, E-water supply; sewerage, waste management and remediation activities. Depending on the needs of users, PPI is expressed in the producer price or in the base price. The producer price is the amount of money obtained by the producer from the buyer per unit of the product (good or service) decreased by the due added value tax on goods and services (VAT) and increased by the excise tax (in the case of the payer of these taxes), taking into account possible surcharges, discounts and rebates provided for in the terms of the contract. The base price is used to ensure harmonization and comparability of data at the international level, including STS statistics within the framework of the European Statistical System. It is defined as the amount of money obtained by the producer from the buyer per unit of the product (good or service) decreased by taxes on the product (VAT and excise duties) and any rebates and discounts, and increased by subsidies on the product. Due to the fact that actual price trends on the market are important for users, the PPI index calculation takes into account all the features of products affecting the price. Their specification must be precise so that in subsequent reference periods the reporting unit can clearly identify the same type of representative of the product or service and can provide appropriate, comparable price information. Products whose prices are compared in subsequent periods should be identical in terms of their technical (physical) and economic (transaction) characteristics. To show the correct trend in price movements, the actual transaction price resulting from the sale/contract is analyzed. The reasons for producer price movements may be of a diverse nature and result, among others from changes in: raw material/semi-finished product costs, labour costs or other operating expenses, taxes or fees, exchange rates, the impact of international oil and gas prices, customer change and/or other conditions of sale, product or service modernization.

Domestic and foreign users benefiting from data on the PPI put attention on their information and deflation functions. Key users of PPI data are following:

International Monetary Fund, European Central Bank, National Bank of Poland – these entities use the PPI information function to shape financial and monetary policy in order to ensure price stability. This is due to the fact that PPI is a preliminary indicator that monitors the emergence of inflationary pressure on the producer’s market, which may subsequently have an impact on the Consumer Price Index (CPI2) and consumer price developments.

Ministries and central offices as well as local authorities that are interested in monitoring of PPI data with reference to the effects of implementing economic strategies and policies and their impact on the whole economy of the country.

Entrepreneurs who assess producer price indices and the economic situation in terms of their impact on their business activities and functioning of other competitive enterprises in the given branches.

Scientists use PPI data for economic analyzes defining the directions of development of various sectors of economic activity and their mutual interactions. For analytical purposes, it is important to apply PPI as a deflator in converting macroeconomic indicators from current prices to constant prices. In this way comparable price conditions are maintained in economic analyses.

Experts in the field of national accounts who benefit from various PPI aggregations (e.g. machinery and equipment producer price index, means of transport procducer price index) and use also this index as a deflator when calculate macroeconomic indicators for the needs of Gross Domestic Product.

The perspective of users oulook on the producer price index and acquired knowlege from their way of thinking have a key impact on improving the methodology and quality of measuring this statistics. Communication with users is a source of inspiration to enrich and develop producer price statistics in terms of greater visibility of position of this statistics in the global and country’s information system3 with reference to:

early warning against business cycle disorders. This information system should also include risk management analyses as the tool used in the formulation of prevention actions against economic crisis;

monitoring the costs and benefits of implementing different economic strategies.

The unmet needs of users are also an opportunity for building a new way of treatment of producer price statistics and can contribute to the development of experimental statistics. Access to various data sources, development of the digital economy and innovations as well as creation of Industry 4.0 require active support of users by statisticians. Statistical staff should act professionally and become guides for users in a diverse world of figures in order to provide them support with regard to access to reliable and good-quality data, and to prevent incorrect interpretations of data. A common challenge for statisticians and users as an opportunity for the amendment of price statistics may become for example the initiative to elaborate the PPI methodological manual in the light of the development of Industry 4.0. which could be undertaken by Eurostat together with the member countries at the level of the European Statistical System.

There are rich documents at international level which give indications and standards for the computation of PPI. The International methodological standards are used to harmonise and standardise elaboration of producer producer price index in industry and to increase its added value for users, namely:

Methodological guidelines of International Monetary Fund and OECD

Producer Price Index Manual, Theory and Practice, International Monetary Fund 2004 [1].

Producer Price Indices-Comparative Methodological Analysis, OECD, Paris 2011 [2].

Methodological guidelines of Eurostat

Methodology of short-term business statistics. Interpretation and guidelines, Eurostat, 2006 [3].

European Price Statistics, An overview, Eurostat, 2008 [4].

Handbook on industrial producer price indices (PPI), Eurostat 2012 [5].

Hovewer, they are not adjusted sufficient to challenges to innovative economy.

3. The methodology followed by Statistics Poland to compute PPI in industry

The requirements arising from relevant European legal regulations4 (available on the website: https://eur-lex.europa.eu/homepage.html) are taken into account in the process of developing methodology of PPI in industry by Statistics Poland.

The source of PPI data in industry in Polish statistics is the monthly report C-01-Report on prices of producers of goods and services for a given month. The subject scope of the survey includes data on average prices of representatives (in the reporting month and in the month preceding the reporting month), as well as the value of sales of representatives. The variables are expressed at producer prices and at basic prices. Data are presented in breakdowns: PPI Poland in total, PPI domestic market, PPI non-domestic market (export). The coverage of the survey is wider than aggregation units for industry and includes reporting units conducting business activity according to PKD07 (equivalent to NACE rev. 2) in the scope of: section A: divisions 02-Forestry and logging and 03-Fishing and aquaculture, Section B-Mining and Quarrying, Section C-Manufacturing, Section D-Electricity, gas, steam and air conditioning supply, Section E-Water supply; sewerage, waste management and remediation activities. PPI in industry is elaborated for grouping NACE rev. 2: B $+$ C $+$ D $+$ E. It refers to divisions of NACE rev.2 for the industry. The sample of units selected for the survey is purposeful and includes about 3200 units and about 30,000 representative items. The selected units constitute 10% of all enterprises with number of employees 10 and more persons. When developing the producer price index in industry, the following classifications are used: Polish classification of activities PKD07, equivalent to NACE rev.2 and Polish classification of products and services PKWiU 2015, equivalent to the Statistical classification of products by activity CPA ver.2.1.

The unit selected for the survey is determined by its significant share in the annual sold production for a given type of activity at the country level. The representatives for the survey are selected by the respondents according to the rules set out by the statisticians of the regional statistical office Opole in cooperation with Statistics Poland -GUS experts. Firstly, these type of activities are selected whose total annual sales value accounts for at least 70% of the total sales value achieved by the reporting unit. 5-digit product groups (categories) according to Polish classification PKWiU are chosen within each selected type of activity. The selection of product groupings of significant sale value in a given type of activity is based on the current and anticipated structure of assortment in production. The sum of the production value of selected categories for which representatives are chosen should constitute at least 60% of the value of the given type of activity. The list of product groupings for which the representative items are selected should be updated during the reporting year if there are significant changes in the production structure within the year, i.e. the introduction of new product groupings into production, whose production share is significant. Within selected categories (5-digit according to PKWiU) specific representatives with a 7-digit symbol PKWiU are selected according to the following criteria: 1) for each grouping of products, representatives with a significant share in the sales value of the representative in the grouping are selected, 2) in the case of production of products in various quality classes, the choice of representatives should be made within individual quality classes, 3) representatives can only be those products that were on sale in the reporting month and should remain in production for a long time, 4) the list of representatives should be updated to reflect changes in structure of the assortment of products in a given grouping, 5) price dynamics of representative item should, to the greatest extent possible, characterize the price changes occurring in the grouping from which it was chosen, 6) the representative items should be named individual products. Representatives may be also unusual products, made to order of a specific recipient, modernized and new products, if their share in the production of a given grouping is significant throughout the year.

In order to obtain price indices for each type of activity by NACE rev.2, they are calculated in the following way:

price indices based on previous month $=$ 100:

for the enterprise level – as weighted average individual price indices of representatives using, as a weight system, the value of their sales in the reference month,

for the level of groups, divisions and sections – as average price indices of lower-level aggregates weighted by their full sales value in the reference month,

price indices on other bases – as average price indices of lower-level aggregates weighted by the value of their sales from 2015, which is updated monthly with changes in prices and changes in the structure of sales occurring in the subsequent months of the year to which the price indices are related.

The monthly producer price index specifying the change in price compared to the previous month is calculated according to the Paasche index based on the weighted arithmetic average using current weights calculated on the basis of current values of sold production of industry for a given reporting period. PPI price indices calculated for various reference periods (e.g. for the same period of the previous year, to December of the previous year, to the base year 2015) are compiled using the Laspeyres chain index and constant weights.

It should be noted that PPI in industry does not take into account qualitative changes [6, 7] in products and is not seasonally adjusted. In Statistics Poland representative items are monitored within the same quality classes. If significant quality changes take place in products, they are treated as new products on condition that their share in the market is significant. Hedonic models are not applied, because they require a lot of assumptions that could be inaccurate in the face of uncertainty, rapid changes in behaviour of producers in economic ecosystem acting differently on different incentives. Respondents in selected enterprises for the survey, having expert knowledge and using the recommendations of Statistics Poland, are responsible for making an appropriate replacement and selecting appropriate, consistent representative products for further price monitoring.

4. Improvement of the quality measurement of the PPI in industry

Users’ needs and their expectations enrich perspective of looking on the PPI in industry with regard to improvements in the whole survey process. They play also crucial role in the quality assessment of PPI. This assessment is based on variety quality dimensions and has the impact on different methodological aspects. Hence, in this section selected methodological issues are presented from the perspective of quality criterias.

One of quality dimension strictly connected with users needs and their satisfaction is data relevance. Taking into consideration that the data relevance is related to satisfying the demand for information, it is worth considering how to maximize the potential arising from this quality component for further development of PPI elaboration. Statistics Poland uses a synergistic approach in the evaluation of the survey process. It means that optimalisation of data relevance results from both improved data accuracy as well as their timeliness and punctuality, accessibility and clarity, comparability and coherence, and ensuring confidentiality and data security. The criteria for quality assessment are interrelated. Results of decisions in the field of quality are visible in different stages of statistical process of PPI. For example, if we introduce additional metadata about the PPI in industry, we can reduce the cases of incorrect interpretation of data. If, for example, as part of accessibility and clarity, we insert in analitical text a chart on which we present trends in shaping the PPI and in sold production of industry as well as in the price index of consumer goods and services, we may enrich the knowledge of users about economic situation and inspire them to acquire new knowledge. Statistics Poland applies such solutions in preparation of the complex publications on prices in the national economy.5

When assessing the accuracy of data in the case of the Polish producer price index in industry, only non-sampling errors are analyzed, because the sample is purposeful. These non-random errors include: coverage errors, data collection errors, measurement errors, processing errors, non- response and assumptions errors. Coverage errors do not occur because reporting units and representative items of goods and services are selected purposefully so as to ensure completeness in collecting price data. The sample survey is updated annually. In the case of occasional liqudation of reporting units and withdrawals of representatives from the market, Statistics Poland replaces them by alternative units and representatives with similar production and sales transaction specifications. As for data collection errors, all information regarding PPI are collected electronically using the Reporting Portal of Statistics Poland, where there is automatic price validation in terms of data completeness, arithmetic and logical relationships between statistical data. The correctness of data is also checked in relation to historical data from previous reporting periods. Employees of the regional statistical office in Opole carefully investigate the causes of producer price changes, in particular when their dynamics are out of the acceptable range of 80%–120%. In the case of data missing on prices of a temporary nature resulting from the seasonality of production, their re-registration takes place after the re-appearance of the product/service on the market. Price data are then imputed backwards by the responding unit. If the lack of price data is permanent, the representative item of the product/service is withdrawn from the sample and replaced by a new representative item with similar production specification, provided that its share in the value of sold production is significant. Statistics Poland eliminates measurement errors of PPI in industry by introducing additional information to the reporting form and to explanation notes as well as in contextual help on the Reporting Portal. Improvements are the result of effective, steady cooperation between statisticians and respondents. Processing errors in Statistics Poland are eliminated at the stage of control tables, which are the basis for analyzing the correctness of algorithms for calculating elementary indices and their further aggregation for the aim of PPI elaboration. The unit non-response rate in the Polish survey is about 5% (the response rate varies within 95%–96%). Assumptions errors do not occur. It is important to notice also that data revisions take place sporadically and are associated with the correction of final data after prior publication of preliminary data. They can also result from changes in the base year.

To ensure users access to the current price trends on the producer’s market, the compliance with timeliness and punctuality of data is of crucial importance for Statistics Poland. The survey regarding PPI is carried out in accordance with the agreed timetable. Preliminary data are published 20 days after the reporting period and final data 50 days after the reporting period. The data dissemination calendar is available on Statistics Poland website: https://stat.gov.pl/en/latest-statistical-news/news-releases/. Publications are available in accordance with the publishing editorial plan on Statistics Poland website: https://stat.gov.pl/en/questions-and-orders/editorial-title-plan-of-the-statisics-poland- and-rso/.

Another quality dimension that includes findings from users needs and their perspective of seeing of statistical information on PPI is accessibility and clarity. Appropriate data accessibility and clarity guarantee users not only easy access to information that they are interested in, but also provide them with the necessary methodological knowledge for the purposes of drawing correct conclusions from own analyzes and preventing so-called “fake news”. For Statistics Poland the most important principle is respecting rule to make data available together with reference metadata e.g. in Knowledge Database Prices. The range of Polish metadata addressed to users in the field of PPI in industry has variety character. Information is available electronically.6

In addition, metadata in English is disseminated by international organizations, e.g. the International Monetary Fund on the website: https://dsbb.imf.org/ sdds/dqaf-base/country/POL/category/PPI00. Microdata is not shared.

In order to ensure comparability and coherence in PPI statistics in industry, Statistics Poland uses international methodological standards of the International Monetary Fund and Eurostat.. The PPI is compiled according to the standard classifications of the European Statistical System ESS: NACE rev.2, CPA rev.2.1. In order to satisfy users needs with regard to ensure comparability over time, Statistics Poland presents long time series over a period of one month from 1996 on the website: https://stat.gov.pl/en/topics/prices-trade/price-indices/price-indices-of-sold-production-of-industry- in-1996-2020,4,1.html. As part of comparability within the domain, preliminary data in Statisics Poland are compared with final data. As part of comparability between domains, data on PPI in industry are compared during validation with other data from enterprise statistics, consumer goods and services price statistics, and from foreign trade statistics in terms of intensity of changes in dynamics of macroeconomic indicators as well as export directions.

Organisational and methodological solutions used by Statistics Poland in the statistical process in the field of PPI in industry are aimed at reducing costs of survey and burden of respondents by such solutions as for example: optimalisation of purposeful sample, electronic data collection, support for respondents at the each stage of carrying out the survey.

The principles of ensuring confidentiality and data security are strictly respected in PPI statistics. They result from the relevant articles of the Polish Public Statistical law (e.g. articles: 10, 38, 38a) [8] and internal procedures of Statisics Poland.

There are sometimes trade offs between quality criteria, e.g. shortening the deadline for publication of the PPI in industry can deteriorate its precision and affects frequent data revision. However, an integrated approach to data quality evaluation including a comprehensive manner can be a source of ideas for improvement in the process of continuous improvements in line with Total Quality Management TQM and with principles of European Statistics Code of Practice [9]. Even the smallest improvement by introducing a new data visualization in the form of a graph, preparing new metadata, developing answers to frequently asked questions by users and respondents can contribute to a durable change in the quality development process of the PPI in industry and can meet users needs.

The another innovative proposals for better satisfying users needs and improving quality of elaboration of PPI in industry is connected with use of “big data” obtained from digital platforms [10]. However, it seems that it is too early to try to replace the analysis of statistical data on producer price indices obtained from the traditional methods and use experimental analysis based on “big data” due to the lack of a harmonized program for assessing the quality of PPI in industry data based on “big data” and lack of operational guidelines on this issue at the level of European Statistical System. It seems that these activities could be coordinated by Eurostat using the present experiences of the Member States.

Given the importance of these issues and also taking into account the specificity of producer price statistics in industry related, among others, with the appropriate selection of units for the survey and with choice of representatives of products/services, it is worth paying attention to the methodological challenges associated with work with “big data”, which are presented in the following table:

Table 1
The scope of activities necessary to initiate the implementation of “big data” when developing the producer price index in industry

Type of action	Scope of work/issues to be identified/methodological dilemmas
Selection of the online platform where electronic commerce is carried out by the producer of goods and services and where transactions of sale and purchase in the field of industry are carried out.	Evaluation of the online platform based on the following criteria: whether the digital platform has local or wider range, whether it concerns one producer of goods and services or groups of producers; whether the scope of data on the electronic platform allows to: identify producer prices in the long period, provide source of prices and analyse the reasons for their changes; does the scope of data include historical data allowing to track the trend in producer prices movements; is coverage of data acceptable for the purposes of PPI calculation.
Selection of NACE divisions significant for a given section (criterion: the value of sale of products and services) in which the experimental test will be performed	The degree of fulfillment of the criteria listed in item 1 a, b, c, d of this table should be taken into account.
Selection of product and service representatives from NACE divisions (see item 2 of this table), where the price index calculated on the basis of web-scraping ${}^{*}$ will be compared with the price index developed in a traditional way based on statistical surveys.	It is advisable to perform a comparative analysis of the calculation of PPI of products and services for a given NACE division on the basis of experimental statistics based on web-scraping with data for the same NACE division obtained from traditional statistical surveys.
Monitoring of information on producer price movements taken from the on-line platform for the same or comparable representatives for a specified period of time.	The following points need to be identified: what period of statistical observation is optimal for the purposes of the analysis; do we have a list of comparable, substitute representative items.
Adaptation of the structure of data on producer prices taken from the on-line platform to the data structure used for statistical purposes.	Work should be undertaken to create a data structure template that will help to further process of “big data” for statistical purposes in the area of producer price statistics in industry.
Evaluation of the quality criteria for “big data” used to calculate PPI in industry.	It is advisable to analyze the different criteria for the evaluation of quality of “big data” taking into account the following aspects: data coverage, their completeness – is all information provided in the analyzed records, are records not duplicated; whether the prices relate to a given reference period; whether it is possible to track the trend over time for a given representative; whether the representatives are not constantly changing, whether there is stability in their observation; whether the conditions for data comparability in subsequent reference periods are met; whether the measurement unit is correctly given/correctly calculated; are identifiable changes in the structure of assortments of products and services.

Type of action

Scope of work/issues to be identified/methodological dilemmas

Selection of the online platform where electronic commerce is carried out by the producer of goods and services and where transactions of sale and purchase in the field of industry are carried out.

Evaluation of the online platform based on the following criteria:

whether the digital platform has local or wider range, whether it concerns one producer of goods and services or groups of producers;

whether the scope of data on the electronic platform allows to: identify producer prices in the long period, provide source of prices and analyse the reasons for their changes;

does the scope of data include historical data allowing to track the trend in producer prices movements;

is coverage of data acceptable for the purposes of PPI calculation.

Selection of NACE divisions significant for a given section (criterion: the value of sale of products and services) in which the experimental test will be performed

The degree of fulfillment of the criteria listed in item 1 a, b, c, d of this table should be taken into account.

Selection of product and service representatives from NACE divisions (see item 2 of this table), where the price index calculated on the basis of web-scraping ${}^{*}$ will be compared with the price index developed in a traditional way based on statistical surveys.

It is advisable to perform a comparative analysis of the calculation of PPI of products and services for a given NACE division on the basis of experimental statistics based on web-scraping with data for the same NACE division obtained from traditional statistical surveys.

Monitoring of information on producer price movements taken from the on-line platform for the same or comparable representatives for a specified period of time.

The following points need to be identified:

what period of statistical observation is optimal for the purposes of the analysis;

do we have a list of comparable, substitute representative items.

Adaptation of the structure of data on producer prices taken from the on-line platform to the data structure used for statistical purposes.

Work should be undertaken to create a data structure template that will help to further process of “big data” for statistical purposes in the area of producer price statistics in industry.

Evaluation of the quality criteria for “big data” used to calculate PPI in industry.

It is advisable to analyze the different criteria for the evaluation of quality of “big data” taking into account the following aspects:

data coverage, their completeness – is all information provided in the analyzed records, are records not duplicated;

whether the prices relate to a given reference period;

whether it is possible to track the trend over time for a given representative;

whether the representatives are not constantly changing, whether there is stability in their observation;

whether the conditions for data comparability in subsequent reference periods are met;

whether the measurement unit is correctly given/correctly calculated;

are identifiable changes in the structure of assortments of products and services.

Source: Own study based on [11, 12, 13]. ${}^{*}$ Data scraping used for extracting data from websites. See more information from the website: https://en.wikipedia.org/wiki/Web_scraping.

To sum up this part of the paper, it is worth considering to take by Eurostat’s initiative to develop a methodological handbook devoted to experimental statistics in the field of PPI in industry on the basis of “big data” for selected NACE divisions of industry. This manual might include elements of risk management in the development and dissemination of producer prices statistics based on “big data”. The above work could be carried out by a group of experts from EU countries as part of an international task group.

5. Methodological challenges and dilemmas in the face of the development of Industry 4.0

The changing world on the one hand causes the occurrence of complex, difficult to predict phenomena, on the other hand creates new cognitive opportunities for further improvement of methodology and quality of data in the field of producer price statistics. In the context of producer prices in industry, the need to explore new methodological possibilities will result, among others, from the short life cycle (exploitation) of goods and services in the innovative economy replaced by new, innovative products, which may hinder the monitoring of price trends due to too short series of data for statistical observations. Another important issue will be the assessment of the role of automation and robotization of Industry 4.0 as a price factor in shaping producer prices and the impact of this phenomenon on methodological solutions in price statistics.

The selection of representatives to examine price changes on the producer’s market in the conditions of technological progress and dynamic innovative changes related to the development of Industry 4.0 [14, 15] as well as the occurrence of unpredictable phenomena, including pandemics and natural disasters, may cause difficulties. A lot of methodological challenges and dilemmas are created. How to provide stable monitoring of producer prices of representative items of products and services in comparable characteristics under such changing conditions? What criteria should be set for selecting representatives in areas where automation and innovation are starting to dominate? Which of the approaches will prove to be more effective in the conditions of dynamically changing reality – whether the development of a central, standard list of representatives for observation of PPI in industry in European countries, or the use of another solution in which representatives are selected by respondents from each country based on significant share in sold production on the market or in turnover? Another issue to consider is the frequency of updating of weights. It is worth considering whether updating the weights every 5 years is sufficient? Should weights be updated more often e.g. every 2–3 years to keep pace with structural changes in economy? Should weights be fixed? Should we use on the larger scale of hybrid weights, i.e. both fixed and changeable? Another problem is related to rapid quality changes of products in the innovative economy. What methods of eliminating the impact of quality changes of products are the most effective to ensure comparability and consistency for the PPI measurement between EU Member States?

Should changes in production costs between the old product and product after a qualitative change be taken into account or should quality changes be measured solely on differences in prices? How to develop a credible PPI in industry based on “big data” from online digital platforms with respecting quality requirements?

6. Concluding remarks

This paper presented a wide and general outlook on improvements in methodology and quality measurement of PPI in industry using some examples from Statistics Poland. Continuous enlargement of knowledge and exchange of experiences in this field are desirable.

On the one hand statistics on PPI in industry is compiled on the basis of sound, international methodological and quality standards. One of them is European Statistics Code of Practice which principles guarantee that adequate institutional environment, statistical processes and statistical output are applied.

On the other hand user needs resulting from the changing economy and their expectations will play an important role in the further development of producer price statistics in industry. They will influence on the improvement of PPI methodology and the quality of the research process. This development will become more and more desirable in the face of changes resulting from the progressive implementation of Industry 4.0. It will “open gates” for new investigation how to effectively construct methodology of PPI in industry in the face of: structural changes and robotization, introduction of innovation, relationships between innovative economy and circular economy i.e. zero waste economy. How will look methodology of PPI in the future with regard to selection of representative items, updating of weights, changes of base years, revision policy, data quality. This methodological work will require constant enrichment of ways of thinking about PPI in the strategic and analitical context, implementation of new instruments and techniques of work. More mutual, coordinated efforts of statisticans in continuous improvements of methodology and quality of PPI in industry are needed on international forums. It seems that activities in good directions in the process of overcoming of methodological challenges would be the preparation of methodological manuals by Eurostat in cooperation with experts of the Member States in the areas of:

experimental statistics based on the development of PPI in industry based on “big data” from online digital platforms. The handbook could contain not only the standard quality assessment criteria for “big data” used to develop PPI in industry but also elements of risk management in the preparation, processing and dissemination of this type of data;

the methodology for developing PPI in industry in the face of the occurrence of extreme phenomena of a global nature and of changes resulting from the development of Industry 4.0.

It is important to notice that producer price statistics not only reflects changes in the economy, but also influences on their shape, so statisticans should strive for continuous improvement of the survey process of PPI beliving that amendmends which are yet unattainable today will become reality in the smart future thanks to technological progress, development of artificial intelligence and obtained knowledge. Users needs are demanding and challenging and statisicans should adequately response giving them indices fit for their purposes. Improvements in methodology and quality measurement of PPI in industry should keep pace with relevance of statistical information. Users play important role in this process as the partner of statisticans.

Footnotes

Short Term Statistics with reference to PPI in industry includes variables: A.310 – Output prices, A.311-Output prices of domestic market, A.312- Output prices of non-domestic market.

For the aim of international comparisons is used HICP- the Harmonised Index of Consumer Prices.

Appreciating the role of official statistics in monitoring of various types of phenomena and counteracting their negative effects for the economy and the society, Eurostat has taken the initiative to create high-value datasets HVD. The variables intended for presentation in the databases include, among others, producer price index in industry PPI.

More detailed information are included in the appendix.

See more details in the appendix to this paper where examples of metadata are presented.

See examples in appendix to this paper.

Acknowledgments

The author expresses gratitude to the reviewers for their comments that contributed to the enrichment of the content of this article.

Appendix

References

Producer Price Index Manual, Theory and Practice, International Monetary Fund 2004. Available from website: https://www.imf.org/en/Publications/Manuals-Guides/Issues/2016/12/30/Producer-Price-Index-Manual-Theory-and-Practice-16966.

Producer Price Indices-Comparative Methodological Analysis, OECD, Paris 2011.Available from website: http://www.oecd.org/sdd/prices-ppp/48370389.pdf.

Methodology of short-term business statistics. Interpretation and guidelines, Eurostat, 2006. Available from website: https://ec.europa.eu/eurostat/documents/1916593/1917176/KS-BG-06-001-EN.pdf/973bf1c8-e598-4228-9d18-2f7704a95e45.

European Price Statistics, An overview, Eurostat 2008. Available from website: https://ec.europa.eu/eurostat/documents/3217494/5700847/KS-70-07-038-EN.PDF/6b78bcc3-b219-4038-80d2-4180a9410678.

Handbook on industrial producer price indices (PPI), Eurostat 2012. Available from website: https://ec.europa.eu/eurostat/web/products-manuals-and-guidelines/-/KS-RA-12-020?inheritRedirect=true.

BLS Handbook of Methods. Chapter 14. Producer Prices Background.2011 Available from the website: https://www.bls.gov/opub/hom/pdf/ppi-20111028.pdf.

BLS PPI Methodology. Available from the website: https://www.bls.gov/ppi/methodology.htm.

The Public Statistical law. Available from the website: https://bip.stat.gov.pl/en/law/law-on-public-statistics/.

European Statistics Code of Practice. Available from the website: https://ec.europa.eu/eurostat/web/quality/european-statistics-code-of-practice.

10.

Radermacher

. Governing by the numbers/Statistical governance: Reflections on the future of official statistics in a digital and globalised socjety. Statistical Journal of the IAOS 35, doi: 10.3233/SII-190562 IOS Press, 2019, 519-537. Available from the website: https://www.researchgate.net/publication/337483376_Governing-by-the-numbersStatistical_governance_Reflections_on_the_future_of_official_statistics_in_a_digital_and_globalised_society.

11.

Auer

Boettcher

. From price collection to price data analytics. How new large data sources require price statisticians to re-think their index compilation procedures. Experiences from web-scraped and scanner data. Statistics Austria 2016. Available from website: https://www.ottawagroup.org/Ottawa/ottawagroup.nsf/4a256353001af3ed4b2562bb00121564/1ab31c25da944ff5ca25822c00757f87/$FILE/From%20price%20collection%20to%20price%20data%20analytics%20-Josef%20Auer,%20Ingolf%20Boettcher%20-Paper.pdf.

12.

Kotzeva

. From Big Data to trusted Smart Statistics. Prezentation Power Point for the meeting BDES in Sofia. 2018. Available from the website: https://webgate.ec.europa.eu/fpfis/mwikis/essnetbigdata/images/a/a9/BDES_2018_Mariana_Kotzeva_From_Big_Data_to_trusted_Smart_Statistics.pdf.

13.

Szutkowska

. Reflections on the future of statisitcs (Polish). Wiadomości Statystyczne. The Polish Statistician, vol. 65, 2020, 2, pages: 43-57 Available from the website: https://ws.stat.gov.pl/Article/2020/2/043-057.

14.

Word Bank. World Development Report 2019: The Changing Nature of Work. Washington 2019. Available from the website: https://www.worldbank.org/en/publication/wdr2019

15.

OECD. The Future of work. OECD Employment Outlook 2019. Available from the website: https://socialprotection.org/fr/discover/publications/future-work-oecd-employment-outlook-2019.