Cities
Single-city data
Comparative data
Production
Production was calculated by combining a number of different sources. There are three parts of production:
CROP
In order to estimate crop production, data were downscaled from production data on national level. In order to estimate London's production, the local agricultural labour force as a share of the larger region was used as a proxy. This strongly affects reliability of these figures. Even if the local production is indeed linked to the labour force, it is unlikely that the crop breakdown follows national patterns. However, it was the closest proxy available.
MEAT
The headcount figures for NUTS2 were used to estimate meat production. An impactful variable is the assumption that the entire headcount (which is the total animal population at a specific moment in time) is indeed slaughtered in that same year. For some animals it might take longer, for others there might be a number of generations slaughtered during the year. Reliability is thus impacted.
FISH
Author estimates fish landings for human consumption to be zero. This might be an undercount.
NOTE: the data quality indicators listed are those corresponding to crop production. Meat production has different DQIs but is expected to be much smaller in quantity, and the crop DQIs were therefore listed.
Data Quality Indicators
| Reliability | 5 |
Non-qualified estimate or unknown origin
The downscaling using labor force and crop production affects reliability (for crops), and for meat the extrapolating from livestock head count and slaughter similarly makes this an indirect estimate. |
|---|---|---|
| Completeness | 2 |
Representative data from a smaller number of sites but for adequate periods
Crop production is relatively complete |
| Temporal correlation | 3 | More than one but no more than three years of difference to year of study |
| Geographical correlation | 4 |
Average date from larger area in which the area under study is included and with which it does not share comparable conditions.
Derived from national data |
| Access | 1 | Publicly and readily available data |
| Frequency | 3 | New data collected every 4-5 years |
| Informality and illegality | 2 | Illegal or informal flows estimated at no more than 5% |
| Classification compatibility | 2 |
Data classified using a different yet compatible classification system
"Permanent crops for human consumption" had to be guessed (now classified as Cereals) |
Imports & Exports
HM Revenue & Customs reports on regional trade. London is one of the regions that is reported on. Trade includes both imports and exports and data are available in using the Standard International Trade Classsification Commodity Code (SITC).
In the Regional Trade Statistics Methodology Paper, data are reported to be derived from Overseas Trade Statistics. Despite an e-mail confirmation from HMCR (Thompson, May 2021) that data include domestic trade as well, the methodology paper makes no clear differentiation between international and domestic trade. However, data on domestic trade must presumably be obtained through different means, as reporting requirements generally vary strongly between local and international trade. Until the exact data collection methodology is fully understood, reliability of this dataset is classified as 'unknown'.
Data Quality Indicators
| Reliability | 5 | Non-qualified estimate or unknown origin |
|---|---|---|
| Completeness | 1 | Representative data from a sufficient sample of sites over an adequate period to even out normal fluctuations |
| Temporal correlation | 1 | Time period is equal to the period of study |
| Geographical correlation | 1 | Data from area under study |
| Access | 1 | Publicly and readily available data |
| Frequency | 1 |
New data collected on an annual basis
Data are in fact available on a quarterly basis |
| Informality and illegality | 2 |
Illegal or informal flows estimated at no more than 5%
We expect a certain level of illegal or informal trade that is not captured in official statistics. |
| Classification compatibility | 3 |
Data classified using a different classification system that is not fully compatible; up to 50% of the data require reclassification
Standard International Trade Classsification Commodity Code is not fully compatible |
Consumption
Three sources were combined:
- Family food datasets - CR - household purchases
This dataset is evaluated for fitness as a proxy for total consumption (minus "eating out" covered in the second dataset). This dataset exclusively covers household purchases, and therefore lacks details on consumption by e.g. the public sector or visiting tourists. Data are being obtained through food diaries. Within the dataset, many of the data points are pro-actively flagged by the data producer as having a high error margin (>20%), affecting the reliability of the dataset. The associated food survey dates 'cover the financial year 2018/19'.
- Family food datasets - CR - eating out purchases
This dataset is evaluated for fitness as a proxy for 'eating out' consumption. Data are being obtained through food diaries. Within the dataset, many of the data points are pro-actively flagged by the data producer as having a high error margin (>20%), affecting the reliability of the dataset. The associated food survey dates 'cover the financial year 2018/19'. Food groupings by type of cuisine (e.g. Indian) and food product (e.g. salads) are different from other datasets, including the household purchases food survey.
- Population estimates for the UK, England and Wales, Scotland and Northern Ireland: mid-2019, using April 2020 local authority district codes
This is the best available and official data on the population estimate for London. However, there are a number of data quality concerns to take into account. Detailed information on the quality is available in the Mid-year population estimates Quality and Methodology Information (QMI). Key limitations, as can be seen there, include the fact that 2019 data is far away from the base census year (2011), and there are quality limitations around internal and international migration. The population does not include daytime populations or short-term visitors (affecting the completeness score). Specific data quality indicators for the mid-2019 population estimate are available (Population estimates: quality information). The rankings in this document indicate a relatively high risk of uncertainty in the estimates for most of London's subdivisions, especially in the "Census base" and "Internal migration" categories.
Data Quality Indicators
| Reliability | 3 | Non-verified data partly based on assumptions |
|---|---|---|
| Completeness | 3 | Representative data from an adequate number of sites but from shorter periods |
| Temporal correlation | 2 | Up to one year of difference to year of study |
| Geographical correlation | 1 | Data from area under study |
| Access | 1 | Publicly and readily available data |
| Frequency | 2 | New data collected every 2-3 years |
| Informality and illegality | 1 |
No illegal or informal flows, or they are fully included
Because the data are based on food diaries, we do not expect "illegal or informal" consumption to be a specific issue of concern here. |
| Classification compatibility | 4 |
Data classified using a different classification system that is not fully compatible; more than 50% of the data require reclassification
These reports take a "product-based" look at food, which means we had to attempt conversion to our "crop-based" classification. Many groups had to be classified under Miscellaneous for lack of a more appropriate category. |
Food loss and waste
We use a report produced by WRAP for the year 2015 called "Courtauld Commitment 2025 - food waste baseline for 2015". The Waste and Resources Action Programme (WRAP) is a registered UK charity operating since 2000 that produces regular reports on food waste and loss. The most recent food loss and waste report across the value chain is from 2015. [methods] It presents food loss and waste data per stage of the value chain in tonnes. We calculate what these figures represent in share of the food supply for that year (using FAOSTAT food balance sheets) in order to be able to extrapolate data for our base year (2019). We then multiply our 2019 london consumption data by the share of food waste and food loss.
Data Quality Indicators
| Reliability | 4 | Qualified estimate (e.g. by industrial expert) |
|---|---|---|
| Completeness | 5 | Representativeness unknown or incomplete data from a smaller number of sites and/or from shorter periods |
| Temporal correlation | 4 | More than three but no more than five years of difference to year of study |
| Geographical correlation | 2 | Average date from larger area in which the area under study is included and with which it shares comparable conditions. |
| Access | 1 | Publicly and readily available data |
| Frequency | 2 | New data collected every 2-3 years |
| Informality and illegality | 1 | No illegal or informal flows, or they are fully included |
| Classification compatibility | 5 | Data classified using a different classification system that is not compatible |