Food Price Changes and Household Welfare: What Do We Learn from Two Different Approaches?

Abstract

The use of a marginal approach can significantly distort the predicted effects of large price variations on monetary welfare over the medium- to longer-term. This paper aims at shedding some light on the differences between a marginal approach and a non-separable agricultural household model with behavioural responses. When behavioural adjustments are allowed, households can adapt their consumption and production patterns by resulting in lower deteriorations in household welfare. The second-order effects introduced in the approach with responses reduce the negative effects due to the first-order consumption effects, with significant differences across quintiles. On average, the second-order effects represent up to roughly 40 per cent of total first-order effects.

Acknowledgements

We are very grateful to John Cockburn, Richard Palmer-Jones and two anonymous referees for useful advice. Luca Tiberti acknowledges the financial support from the Partnership for Economic Policy (PEP), with funding from the Department for International Development (DFID) of the United Kingdom (or UK Aid), and the Government of Canada through the International Development Research Center (IDRC).

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. See, for example, Swinnen and Squicciarini (2012) for a summary of the mixed potential effects of spikes of food prices on food security.

2. As long as a substitute is available in the local market, after a price rise in a given commodity, a consumer might divert immediately towards another good. For example, Ivanic and Martin (2014) make the hypotheses that consumption choices adjust immediately after the price change. On the production side, a household might indeed be constrained by – at least – biological production seasons (seeding and harvesting).

3. Note that the literature commonly uses the concept of net benefit ratio (NBR), which can be defined as the value of net sales (rather than net consumption as originally proposed by Deaton) as a proportion of household consumption.

4. In this study, the wage effects brought about by the consumer price increase is omitted since wage elasticity is likely to be negligible in a context like rural Tanzania (see Loening & Oseni, 2007). Indeed, Ivanic and Martin (2014) estimated the medium-run wage elasticity with respect to the price of maize (which is by and large the main cereal produced by Tanzanian farmers) to be equal to 0.1.

5. A detailed description of the theoretical framework and the estimation strategy is reported in Tiberti and Tiberti (2015). All models’ parameters and variables are fully reported there as well.

6. See also Taylor and Adelman (2003) for a brief but exhaustive analytical description of AHMs.

7. As known, a separable AHM rests on the assumption that hired and family labour inputs are perfect substitutes and that the labour market works perfectly.

8. Non-proportional transaction costs refer to the cost that may change non-proportionally with the traded quantity. This implies non-proportional transaction costs for both (on-farm) labour demand and (off-farm) labour supply, and that both functions are nonlinear (see Henning & Henningsen, 2007b, for more discussion on that).

9. Nose and Yamauchi (2016) estimate long-term changes on production decisions brought about by the 2007/8 food crisis. To do that, the authors use panel data before and after the crisis (that is, from 2007 and 2010). With such data, the authors are able to capture long-term effects, for example, machine investments (fixed inputs) and the substitution effects between labour and capital, and the complementarities between land and capital investments.

10. These authors estimated the welfare effects as the difference of production and consumption effects. Here, we estimate the required compensation which is the difference of consumption and production effects.

11. Cereals prices data are from FAO/Global Information and Early Warning System on Food and Agriculture (GIEWS) online dataset http://www.fao.org/giews/pricetool/ (accessed in February 2015) (FAO, 2015). Given the lack of regional data, we use price changes referring to Dar es Salaam, although we are aware that this may overestimate the effect of price increases. However, the simulation exercise should mainly serve as illustration.

12. For reference, the peak (that is 132% in 2012) in real terms is around 95 per cent.

13. In this paper, with ‘farm households’ we mean those households engaged in cereal production only.

14. Data do not reflect accurately national patterns of near self-sufficiency in cereals, as revealed by FAOSTAT statistics. This is principally due to that fact that TNPS-1 survey is not representative of larger farmers (which are typically net sellers) and, as a consequence, data may underestimate production levels. The survey sampling is based on the population frame, and large farmers are ‘rare events’, which cannot be sampled in a survey of about 3000 households. Overall, it is difficult to compare FAOSTAT statistics with aggregates from household surveys. Indeed, FAOSTAT statistics miss much of the informal cross-border trade.

15. 75 per cent is the cereal price increases occurred in the intermediate period between 2008 and 2012 (that is 2010).

16. This finding suggests a complementary relationship among the products, revealing that an increase in the product price leads new inputs to be drawn into general production. The use of additional inputs, in turn, increases the production of other products (for similar results, see, among others, Fulginiti & Perrin, 1990).

17. This mechanism could have some consequences on nutritional wellbeing. Indeed, diverting consumption towards less expensive commodities may imply that households change diet away from protein/micronutrient rich foods (see Minot, 1998).

18. While interesting, assessing how long the marginal and response effects can take go beyond the scope of this paper.

19. Similarly, D’Souza and Jolliffe (2014) find that the poorest are less affected (in caloric terms) than richer quantiles by food price rise.