j. mernies 12/07/20101 review of the sua exercise strengthening methods/procedures for improving the...

23
J. Mernies 12/07/2010 1 REVIEW OF THE SUA EXERCISE Strengthening methods/procedures for improving the quality of the system outputs

Upload: lenard-summers

Post on 02-Jan-2016

217 views

Category:

Documents


1 download

TRANSCRIPT

J. Mernies 12/07/2010 1

REVIEW OF THE SUA EXERCISE

Strengthening methods/procedures for improving the quality of the system

outputs

J. Mernies 12/07/2010 2

NOTE

This slide show was prepared to support an in person presentation on the subject; used as stand alone product may lead to misinterpretation of purposes and meanings of the slide contents.

J. Mernies 12/07/2010 3

The methodological approach applied included:

• the identification of different types of accounts

• the introduction (or re-introduction) of new elements to the SUA equation system and

•the specification of adequate automatic procedures

J. Mernies 12/07/2010 4

Type of account

The classification is based on :

•the data availability for the different SUA equation elements and •the solutions adopted for estimating the missing data.

J. Mernies 12/07/2010 5

Purpose: account elements reflecting the actual commodity supply and utilization

MAIN ACCOUNTS

1.Data are obtained at least for elements AREA/INPUT, PRODUCTION, IMPORTS and EXPORTS.

2.Data are obtained for element STOCKS or at least for element PUBLIC STOSCKS. In many cases, data for element STOCKS are not available and estimates for this element and the corresponding element CHANGE OF STOCKS, have to be prepared to smooth high and unrealistic variations - even after averaging 3-year periods - of the food estimates.

3.Data and estimates on stocks are relevant for main accounts of some major commodities such as rice, wheat and sugar.

4.Reliable estimates of ratios/rates for elements FEED,SEED, WASTE should be updated regularly.

J. Mernies 12/07/2010 6

Data availability: IMPORTS and EXPORTS only

A value for PRODUCTION is calculated just to take into account a negative trade balance (exports>imports).

PRODUCTION doesn't reflect the actual production of the commodity

If the trade balance is positive, PRODUCTION should be zero and the balance allocated to the appropriate utilization element (e.g.

FOOD). Therefore, the resulting value will not reflect the actual commodity

utilization

Data and estimates on stocks are not relevant in this type of account. In particular, stock variation should not be used to cover

negative trade balances.

TRADE ACCOUNTS

J. Mernies 12/07/2010 7

Data availability: PRODUCTION of derived products only.

PROCESSED is calculated just to take into account INPUTS for derived products

PRODUCTION = PROCESSED

other utilization elements = 0

PRODUCTION doesn't reflect the actual production of the commodity

PROCESING ACCOUNTS

J. Mernies 11/05/2010 8

From the above classification, it is evident that – unless all accounts in the commodity tree are built as main accounts - the values for element FOOD can not be taken as actual food estimates at the single account level but at the overall commodity

tree level only (e.g.: rice and products aggregate).

J. Mernies 11/05/2010 9

Ideally, all accounts should be main accounts. However, in practice main accounts are built for the primary products and some derived products only.

Production data for most of the derived commodities are usually difficult to obtain. However, their trade

may play an important role in increasing (imports) or decreasing (exports) the country food availability.

Trade accounts have the purpose of accounting for these contributions.

J. Mernies 11/05/2010 10

To improve the food estimates it is of central importance to obtain adequate data to built main accounts at least for the

following key type of derived products: flours, oils and alcoholic beverages.

In addition, as food estimates are calculated by difference (balance) in the account equations, it is necessary to obtain data on the production of major non-food products derived from food commodities such as non-food alcohol, alcohol

fuel and bio-diesel and also the amounts of each agricultural input used in these processes.

Of course, this is relevant only in the case of the use of straight agricultural inputs, such as straight vegetable oil to obtain bio-diesel and not in the case of waste inputs, such

as waste cooking oils.

J. Mernies 12/07/2010 11

MULTIPLE INPUT ACCOUNTS (MI)

1. A specific product may be obtained as a result of different transformation processes (e.g.: margarine from different vegetable oils; skim milk as a by-product of extracting butter or ghee).

2. Elements INPUT, EXTRACTION RATE and PRODUCTION of the MI product require decomposition by originating process.

3.The coding system should identify not only the accounts – as in the current FAOSTAT - but also the different processes.

J. Mernies 12/07/2010 12

Joint accounts (J)

1. Two or more products derived from the same transformation process (e.g.: cereal flour and bran).

2. Values of element INPUT of the two (or more) joint accounts should be the same.

J. Mernies 12/07/2010 13

Automatic procedures

• The review of the SUA exercise revealed that manual procedures currently used for inputting values meeting specific conditions, are time consuming and may cause estimation errors.

• Therefore, in the revised accounts presented in the prototype they are replaced by automatic ones.

• These are based on a set of algorithms linking elements within and between accounts.

• An example of automatic imputation is the value allocation for element PRODUCTION to cover negative trade balances in the Trade Accounts.

STRATEGY FOR IMPROVING THE SUA SYSTEM

The conceptual design and main structure of the SUA system should be maintained. The main reference documents are

the ICS User’s Manual, the FAO Commodity Definition and Classification and the Agrostat Codebook.

The information on the transformation processes of the agricultural products has a key role in the measurement of the

food and non-food supplies.   

The list of commodities should not be reduced a priori

• adequate and documented solving equation structures and procedures;

• appropriate distinction between type of accounts implying that the main statistical work is focus on a significantly reduced number of accounts;

• removing manual imputing procedures (time consuming and error generating) and replacing them by automatic ones;

• by concentrating efforts in obtaining the best available production data for the primary and key derived products (such as flours, oils, beverages, fuel from agricultural inputs) and in the appropriate processing and consistency checking of the trade tapes.

The system efficiency and output quality can be improved through

The work flow and team organization can be adjusted to reach a good balance

between the limited human resources available and the expected quality of the

system outputs. 

J. Mernies 12/07/2010 17

Reorganizing the work

Redistribution of responsibilities

Changes to the computerized working system

Collaborative work with countries / national focal points / national counterparts

Establishing and documenting common procedures

J. Mernies 12/07/2010 18

The prototype has some important features:

1. The current manual imputing / estimation procedures - time consuming and potential source of errors – are replaced by automatic ones. Consistently, the additional procedures introduced are also automatic ones.

2. All key parameters and technical factors subject to some variation over time are introduced and coded as SUA elements, allowing direct and transparent access and operation.

3. The effects on the food estimates at the country and commodity tree aggregated levels, of updating or modifying any data or parameter, are automatically shown. These food estimates are shown both as time series tables and trend charts.

A prototype has been developed in line with the proposed approach

J. Mernies 12/07/2010 19

The prototype should be tested by ESS statistical clerks and improved through their feedback and discussions with the responsible ESS officers.

PROPOSAL

The current prototype has been tested in two countries only (in the case of India four commodity trees only). This partial test conducted can provide a first light on

the usefulness (or lack of it) of the proposed approach.

• The prototype containing the equation solving structure and algorithms were developed in an Excel platform.

• The purpose was to illustrate the logic and procedures of the proposal.

• If the approach is considered useful, the next step would be the development of an interactive computer system/programme in a more adequate platform for the first time creation (and required future modifications) of the appropriate equation solving structures for each country in an easier and faster manner.

ISSUES RAISED DURING THE TESTING OF THE PROPOSED PROTOTYPE AND RELATED

ANSWERS

Adopting this prototype it would be necessary to have trade figures as reliable as possible because the production of the derived

products would be based on the export less import relation in the case of derived products

Reliable data on trade for derived commodities having some quantitative importance are needed for obtaining good estimates of the food supply. The adoption of the prototype would not introduce any change in this regard. In other words, if the commodity is not important and we don’t have trade data the corresponding TRADE ACCOUNT is irrelevant and no particular effort in obtaining data is required. On the contrary, if the commodity is important, we should have good trade data, even if the approach proposed in the prototype is not adopted.

Another question is on the Final stocks and the use of stock changes to calculate the trend between years, where do you get the data for Final stocks (or Opening stocks and Closing stocks) and is this data available for how many countries and products?

ISSUES RAISED .....

Two considerations:

1. We should have room in the SUAs to record and use any existing information on stocks. In the case of crops, they usually refer to public stocks only. They are, however, an important information – when available - to take into account.

2. If we are forced to estimate change of stocks as an adjusting element, we should be able to see immediately the effects on the implicit final stocks levels and trends

PENDING ISSUES

Aggregation and standardization procedures:

• Commodity Balances Supply (CBS)• Commodity Balances Demand (CBD)• Food Balance Sheets (FBS)