11-13 May 2011 Working Group meeting “Statistics on Living Conditions” Luxembourg, 11-13 May 2011 Food Consumption patterns through HBS Renata Del Rosario.

11-13 May 2011 Working Group meeting“Statistics on Living Conditions” Luxembourg, 11-13 May 2011

Food Consumption patterns through HBS

Renata Del RosarioAna Martinez Eurostat - Unit F5/ Food safety statistics


2

Index

Background

Food quantities - Data available

Using food quantities from HBS

• Methodology

Next steps


3

Background – Food consumption statistics

Aim: basic statistical data on food consumption in a sustainable and harmonised way (within the ESS)

Work started in 2005: – Task Force to identify

• The needs for data on food consumption• The main determinants for food consumption• Analysis of available data sources

Actions carried out during 2006: – Grants to help the statistical authorities to find efficient ways

to collect data on food consumption (DK, HU, SK, SE, NO) – Call for tenders to design a limited number of indicators to

monitor two of the policy needs: food safety and health and nutrition

Contractor: RIVM


4

Background - Work carried on in 2007

Identification of feasible indicators related to ‘food safety’ (chemical contaminants) and ‘health and nutrition’ policies

Definition of indicators and related variables

Examination of data sources: advantages and disadvantages

Feasibility and technical studies

Guidelines for survey modules to collect missing data


5

Agreed indicators related to health and nutrition

Average consumption (gr. / day) Whole population

Health and nutritional issues Food sources

Energy intake Cereals and potatoes, milk, meat and added fats

Saturated fatty acids Milk and milk products (High/low fat), cheese, meat and meat products (high/low fat), butter, fats (high/low fat) oils

Fish fatty acids Fish and fatty fish

Dietary fibre Bread and breakfast cereals (wholemeal/low fibre), vegetables, potatoes, fruit

Calcium Cheese, milk and other milk products, drinking water

Sodium Bread, other processed cereals, processed meat

Iodine Milk and milk products

Vegetables consumption Vegetables, excluding potatoes

Fruit consumption Fruit and fruit juices

Sugar containing beverages consumption Beverages with added sugar, fruit juices, milk, coffee, tea

Possible data sources: dietary surveys, HBS (DAFNE)


6

Background - Work carried out in 2008

Specific action to obtain the data on food quantities from the Household Budget Survey:

– Written consultation in February 2008: members of “Living conditions” and “Food safety” Working Groups

• Possibilities and willingness to provide the data on food quantities collected already within the HBS

– Requested format for the data transmission and other details specified:

• Micro-data file containing detailed information on quantities at household or at individual level or aggregated table

• COICOP classification of food items• Financial support in the form of Grants proposed

21 countries willing to submit collected food quantities 9 grant agreements signed


7

Food quantities collected from HBSData available at Eurostat

Grants 2008 Without financial help

Micro-data at individual level

RO

Micro-data at household level

BE, BG, GR, LV, LT, HU, MT

EE, ES, FR, PL, RO, SI, SK, FI, UK, NO

Aggregated tables

BE, BG, GR, DE, NL CZ, PT

Data on food quantities collected from 2004/2005 HBS surveys and provided to Eurostat


8

Food quantities - Data available at EurostatReference period

2003: Germany 2004: Spain, The Netherlands, Slovenia 2004/2005: Greece 2005: Belgium, Bulgaria, Czech Republic, Estonia,

France, Latvia, Lithuania, Hungary, Poland, United Kingdom

2005/2006: Portugal 2006: Slovak Republic, Finland 2005-2007: Norway 2007: Romania 2008: Malta


9

Using food quantities collected from HBS

Aggregated tables provided by:

– Czech Republic, Germany, The Netherlands, Portugal

– Belgium, Bulgaria, Greece

Linking micro-data files at household level, for:

– Estonia, Spain, France, Latvia, Lithuania, Hungary,

Romania, Slovenia, Slovak Republic, United Kingdom,

Norway, Belgium, Bulgaria, Greece, Malta, Poland, Finland


10

Food quantities purchased by households– Survey reference year (HA02) – Household ID number (HA04) – Food quantities purchased (HQxxxxx)

Household expenditure– Household ID number (HA04) – Household adult equivalent (HB06.2) – Household size (HB05)– Weights (HA10) – Region (HA08) – Population density (HA09) – Type of household (HB07.1, HB07.2) – Educational level of household head (HC08) – Sex of reference person (HC03)– Age of reference person (HC04)– Activity status of reference person (HC12, HC16) – Occupation ISCO 1988 of reference person (HC18) – Socioeconomic situation of reference person (HC23, HC24) – Net income of household (HH09.9)

Using data on food quantities from HBSLinking micro-data files at household level


11

Using food quantities from HBSMean daily food quantities by adult equivalent

Objective

Calculation by adult equivalent:- Indicators:

• Mean daily consumption of vegetables • Mean daily consumption of fruit • Mean daily consumption of sugar containing beverages

- Mean daily consumption of all COICOP items

- Mean daily consumption of all COICOP items by determinants:• Population density (HA09)• Educational level of household head (HC08)• Region (HA08)• Net income of household (HH09.9) – quartile

Calculation of mean daily consumption of all COICOP items by person

Including Standard Deviation, Max and Min values, Confidence Interval at 95% level


12


Methodology

Step 1: identification of missing values for each household and

each COICOP item; for any HQi = 0, check the corresponding

HEi; if HEi = 0 then HQi = 0; if HEi ≠ 0 then HQi = missing value

For all the countries, except Bulgaria, HQi correspond to

HEiC

For Bulgaria HQi correspond to HEiA


13


Methodology

Step 2: Consumption of vegetables

Calculate the aggregate quantity summing up the corresponding breakdown

Alternative calculation (in some cases, better to avoid "missing quantities" for HQ01176 and HQ01178):

HQ0117 (Vegetables) – HQ01177 (potatoes) + HQ01224


14


Methodology

Step 3: Consumption of fruit

Calculate the aggregate quantity summing up the corresponding breakdown

Alternative calculation:

HQ0116 (fruit) + HQ01223


15


Methodology

Step 4: Consumption of sugar containing beveragesCalculate the aggregate quantity summing up the corresponding breakdown

- version “short”

- version “large”


16


Methodology

Step 5: for each household, for each aggregate calculated (veg,

fruit, sugar) and each COICOP item, mean quantities purchased

by adult equivalent (HQAi) =

where:

HQi = quantity purchased by each Household i of each aggregate calculated and each COICOP item

HB062i = number of adults equivalent in Household i

iHB

HQiHQAi

062


17


Methodology

Step 6: calculation of the “new” weights Wi

where i = ID Household = 1, 2, …., n

n = Total number of Households in the sample

HB062i = Number of adults equivalent in Household i

HA10i = Weight of Household i

Step 7: country average quantities purchased by adult equivalent =

weighted average of values of Step 5

iHAiHBWi 10062

n

in

i

n

in

i

Wi

WiHQAi

iHAiHB

iHAiHBHQAi

1

1

1

1

10062

10062


18


Methodology

Step 8: calculation of mean daily consumption of all COICOP

items by determinants:• Population density (HA09)• Educational level of household head (HC08)• Region (HA08)• Net income of household (HH09.9) – quartile

Step 9: annualize data of Step 7 and 8, if needed

Step 10: daily average of food quantities purchased by adult

equivalent =

values from Step 7 and 8 / 365


19

Using food quantities from HBSMean daily food quantities by person

Methodology

Step 1: for each household and each COICOP item, mean

quantities purchased by person (HQPi) =

where: HQi = quantity purchased by each Household i HB05i = Household i size (number of persons per household)

Step 2: Calculation of the "new" weights Wi‘

where i = ID Household = 1, 2, …., nn = Total number of Households in the sampleHA10i = Weight of Household i

iHB

HQiHQPi

05

iHAiHBWi 1005'


20

Using food quantities from HBSMean daily food quantities by person

Methodology

Step 3: country average quantities purchased by person =

weighted average of values of Step 2

Step 4: annualize data of Step 3 if needed

Step 5: daily average of food quantities purchased by person =

values from Step 2 / 365

n

in

i

n

in

i

Wi

WiHQPi

iHAiHB

iHAiHBHQPi

1

1

1

1

'

'

1005

1005


21

Next steps

Complete data elaboration

Send full results and main conclusions to national

experts of the Household Budget Surveys in the

Members States, for their opinion


22

Thank you for your attention

11-13 May 2011 Working Group meeting “Statistics on Living Conditions” Luxembourg, 11-13 May 2011 Food Consumption patterns through HBS Renata Del Rosario.

Documents

food consumption patterns

food consumption dk

living conditions luxembourg

data transmission

missing data slide

basic statistical data

members of living conditions

microdata file