Coded analytical methods with descriptions & limitations
| Data analysis | Description | Limitations of approach | Assessments identified in |
| Quantitative (unspecified) | Analysis was identified as quantitative (unspecified) when assessments indicated that calculations had taken place but did not disclose the exact method used to produce them. | NA | 1 37 41 45–48 54–58 60 61 63 65–73 75–80 |
| Flows model | A method of analysis that can use multiple data sources to attempt to measure trade flows (the inflows and outflows of cigarettes) between multiple markets in order to estimate consumption. | There is currently no well-established effective flows model approach. KPMG’s ’EU Flows' model and the International Tax & Investment Center’s ’IT flows' model are examples of this approach,19 30 31 33–35 and have been criticised for relying on industry-provided data and methodologically weak estimates.54 | 68 71 80 |
| Tax gap | A tax gap is the difference between the amount of tax that, in theory, should be collected and how much is actually collected. To measure this, an estimate of total tobacco consumption is produced, with legal consumption then being extracted, leaving the ‘gap’, that is, the illicit market.117 | Cannot determine whether illicit cigarettes are counterfeit or contraband (box 1) and cannot distinguish between legal tax avoidance and illegal tax evasion.54 | 71 80 109 |
| Econometric modelling | The use of a mathematical formula, using economic data, which considers the relationship between variables correlated with total consumption (eg, consumer income) and variables positively correlated with ITT (eg, proximity to a jurisdiction with lower price, the level of corruption, etc).51 53 | Requires high-quality (often nationally representative data) and experienced econometricians.54 | 54 59 80 |
| Qualitative analysis | Analysis of non-numerical information such as interviews or focus group outputs. This may involve content, narrative, discourse or framework analysis, as well as grounded theory and ethnographic approaches. | Findings cannot be generalised to larger populations and research quality is heavily dependent on the individual skills of the researcher and their agenda.118 | 45 46 |
| Comparison of export and import statistics | Comparison of reported tobacco exports destined for a country with that country’s reported tobacco imports. Persistent discrepancies between these amounts can indicate large-scale smuggling schemes.51 | Complicated by different countries reporting exports/imports differently (eg, in volume or monetary value) and the timing of the reporting. The trade classification system can also change over time.54 | 46 |
| Unclear | The assessment did not provide enough information to determine the data analytical method/s used in the assessed data. | NA | 37 44 59 62 64 73 74 |
ITT, illicit tobacco trade; NA, not applicable.