Is Local Food More Environmentally Friendly? The GHG Emissions Impacts of Consuming Imported versus Domestically Produced Food

Abstract

With the increased interest in the ‘carbon footprint’ of global economic activities, civil society, governments and the private sector are calling into question the wisdom of transporting food products across continents instead of consuming locally produced food. While the proposition that local consumption will reduce one’s carbon footprint may seem obvious at first glance, this conclusion is not at all clear when one considers that the economic emissions intensity of food production varies widely across regions. In this paper we concentrate on the tradeoff between production and transport emissions reductions by testing the following hypothesis: Substitution of domestic for imported food will reduce the direct and indirect Greenhouse Gas (GHG) emissions associated with consumption. We focus on ruminant livestock since it has the highest emissions intensity across food sectors, but we also consider other food products as well, and alternately perturb the mix of domestic and imported food products by a marginal (equal) amount. We then compare the emissions associated with each of these consumption changes in order to compute a marginal emissions intensity of local food consumption, by country and product. The variations in regional ruminant emissions intensities have profound implications for the food miles debate. While shifting consumption patterns in wealthy countries from imported to domestic livestock products reduces GHG emissions associated with international trade and transport activity, we find that these transport emissions reductions are swamped by changes in global emissions due to differences in GHG emissions intensities of production. Therefore, diverting consumption to local goods only reduces global emissions when undertaken in regions with relatively low emissions intensities. For non-ruminant products, the story is more nuanced. Transport costs are more important in the case of dairy products and vegetable oils. Overall, domestic emissions intensities are the dominant part of the food miles story in about 90 % of the country/commodity cases examined here.

Appendices

Appendix A

1.1 The Linkage of Emissions and Economic Drivers

The following equations are used to link emissions to economic drivers in the model (all variables are expressed as percentage changes in the corresponding levels variables):

$$\begin{aligned} qemo_{jr} =qf_{ijr} \end{aligned}$$

(1)

where, the change in output related emissions in region $r\, qemo_{jr}$ is linked to $qf_{ijr}$, which is the change in the demand for commodity $i$ for use by $j$ in $r$.

$$\begin{aligned} ghgff_{gijr} =qfe_{ijr} \end{aligned}$$

(2)

where, the change in firm’s factor-related emissions in region $r\, ghgff_{gijr}$ is linked to $qfe_{ijr}$, which is the change in the demand for endowment $i$ for use in ind. $j$ in region $r$.

$$\begin{aligned} ghgfd_{gijr} =qfd_{ijr} \end{aligned}$$

(3)

where, the change in emissions from firms’ usage of domestic product $ghgfd_{gijr}$ is linked to $qfd_{ijr}$, which is the change in the domestic good $i$ demanded by industry $j$ in region $r$.

$$\begin{aligned} ghgfm_{gijr} =qfm_{ijr} \end{aligned}$$

(4)

where, the change in emissions from firms’ usage of imports $ghgfm_{gijr}$ is linked to $qfm_{ijr}$, which is the change in the demand for $i$ by industry $j$ in region $r$.

$$\begin{aligned} ghgpd_{gir} =qpd_{ir} \end{aligned}$$

(5)

where, the change in emissions from private consumption of domestic product $ghgpd_{gir}$ is linked to $qpd_{ir}$, which is the change in private household demand for domestic $i$ in region $r$.

$$\begin{aligned} ghgpm_{gir} =qpm_{ir} \end{aligned}$$

(6)

where, the change in emissions from private consumption of imports $ghgpm_{gir}$ is linked to $qpm_{ir}$, which is the change in private household demand for imports of $i$ in region $r$.

1.2 Decomposition of the Change in the Ruminant Products Output

To analyze the change in the output of ruminant products we refer to the linearized market clearing condition for merchandise commodities in GTAP:

$$\begin{aligned} qo_{ir} = SHRDM_{ir}*qds_{ir} + \sum _{s} {SHRXMD_{irs}*qxs_{irs}} \end{aligned}$$

(7)

where:

• $qo_{ir}$ is percent change in industry output of commodity i in region r;

• $qds_{ir}$ is percent change in domestic sales of commodity i in r;

• $qxs_{irs}$ is percent change in exports of commodity i from r to region s;

• $SHRDM_{ir}$ is the share of domestic sales of i in r;

• $SHRXMD_{irs}$ is the share of export sales of i to s in r.

Under scenario EURUMINANTS the decomposition of output in the EU reveals that the dominating portion of the change in ruminant products comes from domestic sales ($-$0.06 %). Next, we decompose the domestic sales using its market clearing condition:

$$\begin{aligned} qds_{ir} = \sum _j {SHRDFM_{ijr} *qfd_{ijr} } + SHRDPM_{ir} *qpd_{ir} + SHRDGM_{ir} *qgd_{ir}\nonumber \\ \end{aligned}$$

(8)

where:

• $qfd_{ijr}$ is percent change in domestic product i demanded by industry j in region r;

• $qpd_{ir}$ is percent change in private household demand for domestic product i in region r;

• $qgd_{ir}$ is percent change in government demand for domestic product i in r;

• $SHRDFM_{ijr}$ is share of domestic product i used by sector j in r at market prices;

• $SHRDPM_{ir}$ is share of domestic product i used by private household in r;

• $SHRDGM_{ir}$ is share of imports of product i used by government in r.

The results show that most of the change in domestic sales of ruminant products in the EU is due to increase in private consumption (0.08 %). To further analyze the factors that cause the increase in the output of ruminant products we decompose its domestic demand using the Eq. (9):

$$\begin{aligned} qfd_{ijs} =qft_{ijs} -ESUBD_i *(pfd_{ijs} -pft_{ijs}) \end{aligned}$$

(9)

where:

• $qft_{ijs}$ is percent change in demand for commodity i for use by j in region s;

• $ESUBD_{i}$ is region-generic elasticity of subst. domestic/imported for all agents;

• $pfd_{ijs}$ is percent change in price index for domestic purchases of commodity i by j in s;

• $pft_{ijs}$ is percent change in firms’ price for commodity i for use by j in s.

The outcome shows that the demand component of ruminant products dominates the substitution effect. Hence, we validate that the change in ruminant products is due to increased domestically sourced ruminant product consumption in the EU.

Appendix B

1.1 Trade Elasticities of Substitution

The following trade elasticities of substitution are used in the model simulations (Table 7).

Table 7 Armington elasticities of substitution for different sectors in the GTAP model

Appendix C

1.1 “Food miles” Experiments for Other Food Products in Different Regions of the Global Model

For the robustness of our analysis of the food miles issue we perform additional “food miles” experiments focusing on changes in consumption patterns of dairy and non-ruminant products, vegetable oils and fats, and processed rice, where available,¹⁰ and then summarize the results in Tables 8, 9, 10 and 11. In particular, we impose taxes on private consumption (while maintaining tax neutrality) of both locally produced and imported dairy and non-ruminant products, vegetable oils and fats, and processed rice, to increase domestic purchases in the region by 50 US$ million, and reduce their import sourced consumption by the same amount.

Table 8 Change in world dairy farms, transport and total emissions under “food miles” experiment in different regions

Table 9 Change in world non-ruminant meat, transport and total emissions under “food miles” experiment in different regions

Table 10 Change in world oilseeds, transport and total emissions under “food miles” experiment in different regions

As can be noted from Tables 8, 9, 10 and 11, substitution of local for imported food products may reduce global GHG emissions due to lower use of international transport and not because of changes in production patterns of these food products. This is especially true for dairy products in the European Union and non-ruminant products in Canada.

Table 11 Change in world paddy rice, transport and total emissions under “food miles” experiment in different regions

Appendix D

See Tables 12, 13, 14, 15, 16, 17, 18, 19, 20 and 21.

Table 12 EURUMINANTS—change in world ruminant meat, transport and total emissions (MTCO2e)

Table 13 CHNRUMINANTS—change in world ruminant meat, transport and total emissions (MTCO2e)

Table 14 EUDAIRYFARMS—change in world dairy farms, transport and total emissions (MTCO2e)

Table 15 CHNDAIRYFARMS—change in world dairy farms, transport and total emissions (MTCO2e)

Table 16 EUNRUMINANTS—change in world non-ruminant meat, transport and total emissions (MTCO2e)

Table 17 CHNRUMINANTS—change in world non-ruminant meat, transport and total emissions (MTCO2e)

Table 18 EUOILSEEDS—change in world oilseeds, transport and total emissions (MTCO2e)

Table 19 CHNOILSEEDS—change in world oilseeds, transport and total emissions (MTCO2e)

Table 20 EUPADDYRICE—change in world paddy rice, transport and total emissions (MTCO2e)

Table 21 CHNPADDYRICE—change in world paddy rice, transport and total emissions (MTCO2e)