Eurostat, the statistical office of the European Union

E1: Agriculture and fisheries

European Commission, Eurostat

L-2920, LUXEMBOURG

The data set comprises the annual sales of active substances contained in plant protection products placed on the national market for one calendar year according to the Regulation (EC) No 1185/2009 (http://data.europa.eu/eli/reg/2009/1185/oj) as amended by Commission Regulation (EU) 2017/269 of 16 February 2017 (http://data.europa.eu/eli/reg/2017/269/oj) as regards the list of active substances.

25 countries have a national legislation on the topic of pesticide sales, while EE, ES, LT and MT do not have a specific national legislation. In DK the national legal act came into force in 1980, in BE in 1994, in RO in 1995 (revised in 2007), in other countries from 2000 onwards (the latter being CY in 2021). Table 2.1.1 displays the list of national legal acts and of the corresponding links by country.

The national legal acts are in line with EU regulation in terms of definitions, variables and unit of measure and establish legal obligation for respondents, the sole exception being FI where there is not such obligation.

Table 2.1.1 - List of national legal acts and of the corresponding links by country

The type of national contact organisation, in charge for the pesticide sales statistics, is different across the countries: national statistical offices (17 countries) and/or national authorities (12 countries) or other national entities (2 countries). Table 2.1.2 displays the name of the national contact organisations and the corresponding type of organisation.

Table 2.1.2 - National Contact organisations and type by country

The classification of the active substances in plant protection products is annexed (Annex III) to Commission Regulation (EU) 2017/269 of 16 February 2017 amending Regulation (EC) No 1185/2009 of the European Parliament and of the Council concerning statistics on pesticides.

The active substances are classified according to chemical and functional similarity. Aggregation levels in descending order comprise: major groups, categories of products, chemical classes and active substances.

The statistics cover the wholesale and retail trade sector operating in placing plant protection products on the market. All authorisation holders of plant protection products report the quantity of all active substances that are placed on the market nationally (incl. import and export data). 

This data collection comprises pesticide sales data based on the term 'placing on the market' as defined in Regulation (EC) No 1107/2009 of the European Parliament and of the Council of 21 October 2009 concerning the placing of plant protection products on the market. The scope of the Regulation covers plant protection products and their active substances as can be seen from the key terms of the act immediately below:

Plant protection products (PPPs): products consisting of or containing active substances, safeness or synergists, and intended for one of the following uses:

• protecting plants or plant products against all harmful organisms or preventing the action of such organisms, unless the main purpose of these products is considered to be for reasons of hygiene rather than for the protection of plants or plant products;
• influencing the life processes of plants, such as substances influencing their growth, other than as a nutrient;
• preserving plant products, in so far as such substances or products are not subject to special Community provisions on preservatives;
• destroying undesired plants or parts of plants, except algae unless the products are applied on soil or water to protect plants;
• checking or preventing undesired growth of plants, except algae.

Active substances: substances or micro-organisms, including viruses, having general or specific action against harmful organisms or on plants, parts of plants or plant products.

Quantity: implies kilograms of the active substances sold.

Regulation (EC) No 1185/2009 of the European Parliament and of the Council establishes a common framework for the systematic production of Community statistics on the placing on the market and agricultural use of those pesticides which are plant protection products as defined in Article 2 (a)(i). The statistics shall apply to the annual amounts of pesticides active substances placed on the market in accordance with Annex I and the quantity of active substances shall be aggregated according to the list in Annex III (as amended).

• Enterprise placing plant protection products on the market;
• Authorisation holder or holder of parallel trade permits.

All enterprises placing plant protection products on the market; authorisation holders and holders of parallel trade permits.

The entire territory of each single country, thus the reference area covers the EU, Iceland and Norway.

The following geographical coverage specificities were reported in the reference year 2019 by the following Member States:

- DK: Faeroe Islands and Greenland (included)

- DE: Büsingen am Hochrhein and Helgoland (included)

- EL: Mount Athos region (included)

- ES: Canary Islands, the Balearic Islands, Ceuta and Melilla (included)

- FR: Overseas departments (included)

- IT: Livigno and Campione d'Italia (included); Vatican and San Marino (not included)

- NL: Aruba, Curaçao, Sint Maarten, Bonaire, Sint Eustatius and Saba (not included)

- PT: Azores and Madeira (included)

- SE: Åland Islands (included)

The current legislation is applicable since reference year 2011. Data for HR has been available since 2013, data for IS since 2015.

This report refers to the data collection on pesticide sales statistics for reference year 2019.

Not applicable for pesticide sales statistics, because it is not based on an index number of time series.

Compilation of pesticide sales statistics is based on multiple sources: administrative data (17 countries), surveys (4 countries), censuses (6 countries); 2 countries report a combination of these. Table 3.1 displays the information by country about the type of data source, the data source and the type of data.

Table 3.1 – Type of data source, data source and type of data by country

Annual.

Data collection methods vary a lot among countries as it is displayed in Table 3.3. Respondents are mainly requested to report the data by themselves, either on-line or electronically; in EL data are automatically registered in a national database each time a pesticide is sold to final user while in RO data are collected via a face-to-face interview.

Table 3.3 - Data collection method per country

To ensure statistical data quality, data validation procedures are in place in all countries, as shown in Table 3.4. Consistency of data over time, completeness, outliers and consistency of aggregates are the most common validation checks. 15 countries come back to respondents when inconsistency or any another anomaly is detected. In addition to the validation procedures listed in Table 3.4, some countries perform other checks; BG, in case of deviations, proceeds with confirmation procedure; DE performs plausibility checks; ES performs the quality control of the CATI data recording; LV checks the reliability of the data and LT performs the primary data entry control.

Table 3.4 - Data validation by validation procedure

Pesticide sales data are collected with reference to the weight of the commercial products available on the market. Thus, data are transformed, mainly by using converting tables, in kg of active substances. Aggregates are then created by summing up the active substances according to Annex III of Regulation (EC) No 1185/2009. 

Not applicable for pesticide sales statistics, because the data collection is not based on time series.

Quality of pesticide sales statistics is assured trough the quality of the administrative data source (full survey/census), combined with the validation procedures and by following the quality policy as established in national and European laws and codes of practice.

Quality management covers systems and frameworks in place to manage the quality of statistical products and processes.

In general, countries assess the quality level of pesticide sales as good and satisfactory due to:

• Full surveys/censuses/administrative data (no sampling error),
• Mandatory response for respondents (and thus missing values are absent and no need for imputation),
• Absence of estimation,
• Coherence,
• Accuracy, and
• If necessary, respondents are contacted back for clarification.

11 countries (DE, EL, IT, CY, LT, MT, NL, PL, PT, SK, SE) refer to a general quality policy (generally in line with ESS Code of Practice and/or ESS Quality Assurance framework; DE, LT, SK have also ISO certifications). 

Pesticide sales statistics are of interest for several groups of users that seek data for different reasons.

At national level, countries report that users are represented by:

• National policy makers (e.g. Parliament, Ministry of Agriculture, Ministry of Environment, Ministry of Health),
• National Authorities (e.g. Food Safety Agency, Environment protection Agency, etc.),
• Universities and research institutes and students,
• Consumer organisations and other non-governmental organisations,
• Lobbies of pesticides producers and of farmers,
• General public and media, and
• Business operators.

Pesticide sales statistics data are mainly used by users to:

• Assess the possible risks of using pesticides from an environmental point of view and from public health perspective,
• Measure the progress achieved in the reduction of risk from the use of plant protection products,
• Calculate the harmonised risk indicators,
• Evaluate the trend of the use of pesticides over time,
• Carry out post-registration monitoring tasks of the authorisations (both of PPPs and of holders),
• Assess the goals of national strategy on pesticide use, and
• Analyse market shares.

At international level, data users are the EU Commission and International Organisations (e.g. FAO).

4 countries refer to unmet user needs:

- DE, PT and SI refer that there is a national request by some users for regional data.

- NL and SI report there is a national request for a lower level of aggregation of data (categories of products, chemical classes and active substances).

Detailed information for each country on the main national and international users can be found in the national quality reports.

No user satisfaction survey/assessment is carried out specifically for pesticide sales statistics in any of the Member States.

However, EL, IT, CY and LT carry out general user need satisfaction survey; EE, ES and HU hold regular meetings with main data users and stakeholders (both private users and public users); DE and NL report that since 2019 pesticide data (at level of active substances) have not been confidential any longer by law, allowing for meeting the request of data users for more detailed information on pesticides. 

The data collection covers all active substances of plant protection products placed on the national market and the information of all authorisation holders.

All countries, except EL, reported a 100% data completeness rate. For EL this rate is at 95%.

The analysis of the accuracy and reliability considers the analysis of the sampling and non-sampling errors.

Pesticide sales statistics are largely based on administrative data and censuses/full surveys, therefore sampling error is not influent.

Non-sampling errors are further broken down into coverage errors, measurement errors and non-response errors.

Countries deem missing values as not influent as there is largely a legal obligation on the respondents to provide the data.

Not applicable since data is collected either from administrative registers or from full surveys.

Not applicable.

See sub-points 6.3.1, 6.3.2, 6.3.3, 6.3.4 and 6.3.5.

A coverage error is due to a divergence between the survey population and the target population resulting from an imperfect frame. In the case of pesticide sales, countries report that coverage errors are unlikely to happen as data come from either from administrative registers or from full surveys and there is legal obligation on respondents to provide the information. 

In general, it is not applicable since data are either from administrative sources or from censuses/full surveys. However, PL refers 2.4% of over-coverage rate (due to 3 statistical units either in liquidation or inactive) and RO 0.7%.

Not applicable for pesticide sales statistics as the data stem from one source.

The measurement error is an error that occurs during data collection and causes the recorded values of variables to be different from the true ones. Recording of wrong values can be linked to survey instruments (form, questionnaire), respondents or interviewers.

In the case of pesticide sales, some countries refer to measurement errors due to:

i) Incorrect measurement units (CZ, DE, EE, LU, RO)

ii) Issues with the list of PPPs (outdated, incorrect name/code) (CZ, ES, LU, RO, FI)

iii) Mistyping errors (EL)

iv) Duplications (ES)

Countries adopt actions to reduce the measurement errors: e.g. consistency checks against results of previous years, automatic in-built validation processes and data-processing systems, detailed instructions for respondents on how to fill in the reporting forms, possible re-contacts of respondents (refer to point 3.4 for more detailed information).

In general, due to the use of either administrative registers or full surveys and due to legal obligation on respondents to provide the information, the non-response error is unlikely to happen.

In most countries, the unit non-response rate is either not applicable (due to either administrative data or full surveys and/or legal obligation for respondents) or negligible (e.g. thanks to reminders sent to respondents).

7 countries report the non-response rate as follows:

- DE: overall unweighted non-response rate was 3.5% (due to 0.7% unweighted non-response rate for authorisation holders and 8.1% for holders of parallel trade permits)

- IE: 10% of non-respondents accounting for 5% of sales of active substances

- ES: 12.8% (due to 26.1% of foreign authorisation holders who did not reply)

- IT: 28.38%

- LU: 18%

- PL: 34%

- RO: 3.5%

In general, countries assess the item non-response rate to be not relevant (zero/negligible/not applicable/not available/not calculated).

DE calculates the item non-response rate on 3 different totals:

i) Overall item non-response rate for 2019 was 0.49% (no response for 18 of 3707 PPPs for which a report was expected).

ii) Item non-response rate based only on the 2698 parallel trade products for which a permit was valid in 2019 was 0.48%.

iii) Item non-response rate based only on the 1009 products with a regular national authorisation was 0.5%.

Countries indicate as possible sources of processing error: data entry, wrong code assignment to active substances, unit of measurement.

To avoid/minimise those errors, countries have in place validation procedures, as described in point 3.4 above, both on primary data and on calculated aggregates. If errors are detected, they are corrected base on the comparison with data from the previous years and often by contacting back the respondents.

Not applicable as data are from either administrative sources or from censuses/full surveys and there is a legal obligation on respondents. 

Not applicable for pesticide sales statistics, because the data collection is not based on estimations.

Not applicable for pesticide sales statistics, because the reported data on plant protection products cover the whole year.

Pesticide sales statistics are annual, and data as published at national level and as transmitted to Eurostat are final and thus not subject to scheduled revisions. Countries may revise data, according to either national revision policy and/or the ESS Code of Practice, if necessary (e.g. a data provider/respondent informs about an error made). Revised data are then immediately transmitted to Eurostat and published at national level with an explanatory note (either a footnote, or a flag or an information notice). 

Data revision may be done only if necessary.

Only 4 countries report data revisions.

- DK reports to make a single correction per year.

- DE reports to do 1 or 2 revisions per year.

- IE reports to have revised 2019 dataset, which has been transmitted to Eurostat and published on DAFM website.

- NO reports that the size of yearly adjustments is usually less than 10 litres or kg of pesticides.

According to the Regulation (EC) No 1185/2009, the data is published by Eurostat 15 months after the end of the reference year and the national quality report 20 months after the end of the reference year. 

Not applicable for pesticide sales statistics as the data and national quality reports are only published once as final.

However, PL reports that at national level, the data on PPPs sales in natural units (in commodity mass), broken down into main groups are disseminated as first results with the flag ‘preliminary data’. Preliminary data for 2019 were published on the 06th of July 2020.

Most countries refer to the time framework set by the Regulation (EC) No 1185/2009. 5 countries have released the data prior of the deadline:

- EE disseminates the data 6 months after the end of reference year.

- EL and CY disseminate the data 12 months after the end of the reference period.

- IT reports that 2019 data at national level were released 386 days after the end of the reference period.

- SI releases the data 9 months after the end of the reference period.

Deadline for transmission to Eurostat are set by Regulation (EC) No 1185/2009 (i.e. data must be transmitted until 30 December of the year t+1, and national quality reports until 31 March of the year t+2).

The vast majority of countries transmitted and published data and quality report on time.

5 countries state that a delay occurred in data transmission to Eurostat and 3 countries in quality report transmission to Eurostat as per reference year 2019:

Data transmission:

- DK: 76 days of delay due to underway digitalisation of the process.

- LU: 4 months of delay due to transition year. The responsibility of pesticides statistics was transmitted from the national statistical office (STATEC) to SER.

- NL: 30 days of delay. The data source NVWA delivered data with delay.

- AT: Delay due to high workload.

- NO: Delay due to an error.

Quality report transmission:

- LV: 1 day of delay due to delays in reconciling information.

- LU: 4 months of delay due to transition year. The responsibility of pesticides statistics was transmitted from the national statistical office (STATEC) to SER.

- PT: 13 days of delay due to simultaneous time constraints of IFS and pesticide use data collection.

Data are collected on a country level (NUTS 0). 

Not applicable, because there are no mirror flows in pesticide sales statistics.

Pesticide sales statistics first data collection, as set by Regulation (EC) No 1185/2009, was in 2011. Every 5 years, the classification in Annex III of Regulation (EC) No 1185/2009 is revised. Some active substances are then allocated to groups different from the previous 5-year-period.

In 15 countries pesticide sales statistics data are available since 2011 (BE, CZ, DK, DE, EE, IE, FR, IT, CY, LT, MT, NL, AT, RO, SE) and thus the length of time series is 9 years.

FR indicates that between 2011 and 2013 occurred a change of the categories of distributors of plant protection products subject to the compulsory declaration but still the time series is available from 2011.

9 countries (BG, ES, LV, LU, HU, PT, SI, SK, FI) refer to not have reclassified yet the active substances after the last reclassification exercise and therefore there is a break in time series in 2016. Therefore, the time series is broken in two parts: 2011-2015 (time series length: 5 years) and 2016-2019 (time series length: 4 years).

5 countries have a time series starting in a different year, as displayed in Table 8.2 (table listing countries with their unique first year of comparability and their length of time series):

EL indicates 2014 as first year of comparability due to the low response rate in the previous years and a change in the data collection methodology applied since that year.

PL and NO have longer time series, while HR and IS started later than 2011 to collect pesticide sales statistics.

Table 8.2 - List per country with unique start of comparability and length of time series

See 8.2.

Cross domain coherence refers to the consistency of outputs produced by different statistical processes within the country. Pesticide sales statistics might be compared with pesticide use statistics and other relevant national data collections (e.g. trade, internal production of manufactured goods). However, full comparison and consistency is not possible due to different reference periods and/or differences in methodology (e.g. unit of measure, group composition, statistical concept, etc.). Therefore, only 9 countries (BE, BG, CZ, EE, IE, CY, NL, PL, SK) state to make comparison with pesticide use statistics (when same year is available) even if some caveats are in place (e.g. the pesticide sales statistics include PPPs for different sectors, not only for agriculture).

FR reports that pesticide sales statistics are consistent with those provided by the Union of Industries for the Protection of Plants (UIPP).

Not applicable for pesticide sales statistics, because the data collection is annual.

Not applicable, because it is not related to national accounts. However, SE refers that pesticide sales statistics are reconcilable with National Accounts in the national statistics ‘Quantities of pesticides sold’ (https://www.kemi.se/en/statistics/quantities-of-sold-pesticides).

The large majority of the countries report that the pesticide sales statistics data are internally coherent since they are from one data source only.

IE, FR and AT assess the internal coherence of pesticide sales statistics data by comparison with another data source (IE, Fee register; FR, UIPP data base; AT, Green Report of the Federal Ministry of Agriculture, Regions and Tourism).

4 countries publish a news release on their website related to pesticide sale data. Table 9.1 displays the countries and links to the news related to 2019 data.

Table 9.1 - News release link by country

14 countries make available an online publication on their website. Table 9.2 displays the links to the on-line publications by country.

Table 9.2 - Publication link by country

9 countries have an online database that users can consult and extract data (Table 9.3.a).

8 countries make available predefined data files (.xls, .csv, .pdf) for downloading (Table 9.3.b).

Table 9.3.a - Online database link by country

Table 9.3.b - Downloadable data file

Not applicable or not available. Only IT reports 4 189 consultations in 2019.

Countries have different approach to microdata access. 22 countries may grant access to microdata at level of active substances upon a specific request (BE, BG, CZ, DE, EE, IE, ES, FR, HR, IT, LV, LT, LU, MT, PL, RO, SI, SK, FI, SE, IS, NO).

DE specifies that sales data of individual plant protection products is considered confidential business information and may not be made available to the public without previous agreement from the authorisation holder.

NL specifies that in 2020 data of sales from 2010-2019 have become available to the public. However, microdata at the level of authorisations or authorisation holder of sales are not accessible (e.g. for researchers).

PT specifies that microdata on individuals and enterprises shall only be supplied for scientific purposes, in anonymised form to accredited researchers.

5 countries do not give access to microdata (DK, EL, CY, HU, AT).

No other dissemination formats.

13 countries complement data with methodological documentation, as displayed in Table 9.6.

Table 9.6 - Link to documentation on methodology by country

For most countries (22) there is no other documentation on quality management than the national quality report transmitted to Eurostat. 

DE and IS have quality documentation for internal use only.

4 countries publish online the quality documentation:

- FR publishes a methodological document accompanying each online data publication:

• https://agreste.agriculture.gouv.fr/agreste-web/disaron/Chd2104/detail/
• https://ssm-ecologie.shinyapps.io/Cartes_phytos_BNVD/
• https://www.statistiques.developpement-durable.gouv.fr/plan-de-sortie-du-glyphosate-etat-des-lieux-des-ventes-et-des-achats-en-france-en-2019
• http://www.data.eaufrance.fr/jdd/bd45f801-45f7-4f8c-b128-a1af3ea2aa3e

- IT publishes a methodological document regarding the survey on plant protection products provided for agricultural use (http://siqual.istat.it/SIQual/visualizza.do?id=5000024&refresh=true&language=EN).

- LT publishes quality documentation at http://osp.stat.gov.lt/en/metainformacija3 (available in Lithuanian and English).

- MT publishes quality documentation at https://metadata.nso.gov.mt/reports.aspx?id=16%20.

Not requested.

Not requested. However, IT reports that the national quality report in Italian (https://www.istat.it/it/archivio/216611) has about 500 consultations per year.

Administrative cost and burden on respondents for compiling pesticide sales statistics vary among countries.

9 countries (BG, CZ, FR, HR, CY, LU, AT, PT, NO) report that the information on cost and burden is not available.

7 countries (BE, EL, LV, MT, NL, RO, SE) refer that no extra cost is needed, and the burden is minimal.

6 countries (DK, DE, EE, IE, ES, IT) report a calculated cost laying in the range: 11 900€ (EE) - 77 500€ (DK).

7 countries assess the burden in hours or Euros: DK (1-7 hrs), EE (3 hrs), ES (1 hrs), LT (3.5 hrs), PL (2 hrs), SK (5 hrs), DE (109 650€).

6 countries estimate the administrative labor input: HU (10 man days), PL (1 433 hrs), SI (0.25 man/year), SK (4 people involved per year), FI (19 days), IS (unknown for the Environmental agency, 3-4 days at Statistics Iceland for data processing and sending).

Regulation (EC) No 1185/2009 concerning statistics on pesticides in Article 3 (4), establishes that ‘for reasons of confidentiality, the Commission (Eurostat) shall aggregate the data before publication in accordance with the chemical classes or categories of products indicated in Annex III, taking due account of the protection of confidential data at the level of individual Member State’.

6 countries refer that at national level data on active substances are not considered confidential (CZ, DK, DE, FR, NL, IS). FI reports that data on active substances are considered confidential only if they could be associated with a single producer/authorisation holder and used to deduce the product sale volumes of this individual producer/authorisation holder. SE reports that in case of data on active substances that can be attributed to one or two companies only, the data can be disclosed only by written consent from the authorisation holder(s) concerned.

Individual data (i.e. single producer/authorisation holder) are always treated as strictly confidential by all countries.

All countries strictly follow Regulation (EC) No 223/2009 of the European Parliament and the Council of 11 March 2009 on European statistics as complemented by national acts (if available) on confidentiality policy. 8 countries (DE, EL, ES, LU, MT, PL, RO, SK) explicitly refer to legal obligation (by written declaration/oath) for the staff having direct access to individual and personal data, to ensure full protection of confidentiality without exceptions.

Data deemed as confidential are not published. 17 countries state to apply one or more rules to guarantee confidentiality.

Primary confidentiality

• threshold rule: cell is confidential if less than 3 entities contribute to the total of that cell
• dominance rule: the cell is confidential if the largest unit contributes more than 90% to the cell total (ES has set the threshold to 85%)

Secondary confidentiality

• group level data is declared confidential if it contains 1 - 2 confidential subunits
• group level data is declared confidential if it is necessary for dissemination of the parent group

DE, IT and PT explicitly refer to a secure data storage, with limited access and/or anonymised dataset (i.e. data are stored with no direct reference to the respondent unit). 

The majority of countries do not have further comments.

HU refers a caveat about possible item under-coverage due to some plant protection products authorised as soil improvers and not as pesticides. Therefore, the sales of those products are not collected as there is no obligation.

FI adds as remark that the national sales statistics are dominated (~70%) by a single active substance, urea (Fungicides and bactericides), which is used solely in forestry. This should be accounted for when calculating sales statistics per utilised agricultural area (kg/ha).