What drives unrecorded production and consumption?

Quantitative modeling

Quantitative modeling

Often, the preferred approach to examining the predicted impact of policies ― particularly pricing and taxation policies ― is quantitative modeling.

The starting point of any modeling exercise on reactions to policy changes based on price is to collect data on the price of the duty-paid product and the unrecorded segments, as well as data on distributors’ and retailers’ margins. The aim is to collect data on prices at various points of the alcohol supply chain, particularly those at which taxes are applied. This allows for modeling of the impact of a particular tax change on the retail selling price the consumer pays.   

In practice, because of limited data availability, most modeling exercises begin with data on the retail price of the duty-paid product, broken down into its various components. These include the pre-tax price of the product, the current rate of excise duty, and the VAT. In most countries, the price is likely to vary for the same product across the off-trade and on-trade distribution channels. In other countries, tax rates vary across these channels, as well. Because of these differences, it is ideal to collect data for off- and on-trade distribution channels for each type of drink.

Retail price data on the duty-paid market can be obtained from a variety of sources, including:

• market research firms;
• alcohol trade bodies
• store visits; and
• websites of on- and off-trade firms.

The price of cross-border personal imports can be obtained through this same approach, collated for the countries from which the imports originate.

Data on prices in the unrecorded market are more difficult to ascertain. Because sellers in the unrecorded market tend not to advertise prices openly, prices can only be obtained by someone making test purchases or by asking consumers directly through surveys. The appropriateness of the former approach will vary from country to country depending on the laws in place locally.

Once price data is collected, it is necessary to calculate how much of a duty change is passed through to the pre-VAT price. This may vary across distribution channels. For example, in some countries some supermarkets may absorb, or do not fully pass through, tax increases and instead use alcohol as a loss leader for sales of other products. Conversely, in on-trade channels, it is often argued that “over pass-through” occurs, as publicans and hoteliers use the timing of an excise duty change to increase their prices. Pass-through rates may also vary across product categories and over time.

A number of options exist for modeling pass-through rates. Deciding which is most appropriate will depend on the data available, knowledge of the market, and the extent to which the issue has been previously analyzed in that market.

The simplest approach is to assume full pass-through and undertake sensitivity analysis around it to determine if the issue is important.^[1] A second option is to take the findings of an academic study on pass-through rates.^[2] And a third is to undertake a regression of the price of the drink type on constituent costs (e.g., unit wage costs and input costs) and excise duty in order to estimate the relationship between the variables (see Illustrative Example F).

Whatever the approach taken to modeling the pass-through of changes in excise duty to the retail price, the new retail prices obtained must be combined with those of unrecorded products, including products purchased through cross-border shopping, to obtain an estimate of the new total market price after the duty change. A complicating factor is that little is known about how unrecorded prices change in response to excise changes (or even what initial unrecorded prices are, for that matter), although it is assumed that the cross-border prices of personal imports do not change as a result of the duty changes.

The analysis requires two estimates of the responsiveness of sales volumes to changes in price (i.e., price elasticities) for duty-paid and total consumption. By combining these price elasticities and the projected price changes after a duty change, it is possible to forecast total consumption of alcohol and of the duty-paid segment. This can then be subtracted from the base forecast (the projected consumption absent a tax / price increase) to investigate the impact of the policy change on the amount of duty paid and the combination of cross-border and unrecorded products that are consumed.

Potential sources of price elasticity estimates include:

• ministries responsible for tax collection;
• research papers (by academics, alcohol trade bodies, etc.), including original empirical studies and reviews of the literature (e.g., Gallet, 2007; Wagenaar et al., 2009); and
• an original regression analysis.

The main challenge encountered when attempting to estimate the impact of a price or tax policy on the unrecorded market is that price elasticity estimates are often very difficult to obtain for the total (recorded and unrecorded) market. In some cases it may be possible to derive elasticity estimates from consumer surveys (see previous page - What Drives Unrecorded Production and Consumption: Population-Based Surveys).

Typically, to discern the wider economic impact of a price (or availability) policy, the percentage change in duty-paid consumption resulting from the policy is applied to the results of an economic impact study (see What is the Revenue Loss for Governments). This method has its weaknesses because some of the economic activity in the industry itself, its purchase of inputs, or wage expenditures may not vary with production levels.

While this analysis allows the investigator to estimate the effect of a price or availability-based policy change on the activity the duty-paid alcohol industry supports throughout the economy, it will not show how the policy affects the economic impact of different parts of the unrecorded market. For example, it will not reflect the boost to the transport and travel industries resulting from a policy that increases the demand for personal imports of cross-border alcohol (either through relative price changes or increases in personal allowances).

Harder to quantify are some of the other economic costs associated with duty changes that impact the size of the unrecorded market. Such economic costs that are not readily amenable to measurement may include:

• medical costs from the damage to health from the consumption of unrecorded alcohol;
• decreased investment incentives for legal alcohol manufacturers and retailers; and
• diminished value of manufacturers’ brands and intellectual property due to counterfeit products

[1] Full pass-through is consistent with the market being perfectly competitive. Under or over pass-through is consistent with sellers having some market power.

[2] For a range of academic studies on pass-through rates for alcohol see the references in Ally et al. (2014).