Backcasting United Nations Statistics Division Overview o Any

Backcasting United Nations Statistics Division

Overview o Any change in classifications creates a break in time series, since they are suddenly based on differently formed categories o Backcasting is a process to describe data collected before the “break” in terms of the new classification

Overview o There is no single “best method” o Factors influencing a decision include: n type of statistical series that requires backcasting n n n n (raw data, aggregates, indices, growth rates, . . . ) statistical domain of the time series availability of micro-data availability of "dual coded" micro-data (i. e. businesses are classified according to both the old and the new classification) length of the "dual coded" period frequency of the existing time series required level of detail of the backcast series cost / resource considerations

Main methods o “Micro-data approach” (re-working of individual data) o “Macro-data approach” (proportional approach) o Hybrids thereof

Micro-data approach o Consists of assigning a new activity code (= new classification) to all units in every period in the past (as far back as backcasting is desired) n No other change is required n Statistics are then compiled by standard aggregation n Census vs. survey (weight adjustment issue)

Micro-data approach o Requires detailed information from past periods (for all units to be recoded) n More detailed than just the old code o If information is available, results are more reliable than those from macroapproaches

Micro-data approach o Issues: n Resource intensive n Need solutions if unit information is not available for a period (not collected, not responded) o Nearest neighbor vs. transition matrix approach

Macro-data approach o Also called “proportional method” o This method calculates a ratio (“proportion”, “conversion coefficients”) in a fixed dual coding period that is then applied to all previous periods o The ratios are calculated at the macro level n Could be based on number of units only o Low resource approach o Has a more approximate character

Macro-data approach o In simple form, applies growth rates of former time series to the revised level for the whole historical period o More sophisticated methods may use adjustments based on experts’ knowledge n Example: mobile phones

Macro-data approach o Assumes that the same set of coefficients applies to all periods n This means it is assumed that the distribution of the variable of interest has not changed between the old and the new classification o Applied to aggregates; does not consider micro-data o Relatively simple and cheap to implement

Macro-data approach o Steps: n 1 – estimation of conversion coefficients o Done for dual-coding period n n Longer/multiple periods help in overcoming “infant problems’ of the new classification and allow for correction of data Based on selection of specific variable n 2 – calculation of aggregates using the conversion coefficients o Weighted linear combination n 3 – linking the different segments o Old – overlap – new series o Breaks caused by mainly by change in field of observation n Simple factor or “wedging” n 4 – final adjustment o Seasonal etc.

Comparison o Micro-data approach better retains structural evolution of the economy o Micro-data approach does not require choice of a special variable o Macro-data approach reflects evolution based on fixed ratio for a fixed variable n Seasonal patterns may be distorted o Macro-data approach is more cost-efficient n No consideration of micro-data necessary o Assumptions underlying the macro-data approach become invalid over longer periods n “Benchmark years” might help to measure the effect, if data is available

Other options o Combinations of both approaches are possible n Ratios for the macro-data approach could be calculated for shorter periods only n Micro-data approach could be used for specific years and the macro-data approach for interpolation between these years o Many factors can influence the choice (see beginning) but data availability is a key practical factor