By Joern Fischer
Scale has fascinated geographers for a very long time – but still, this important concept is not getting the attention it deserves by sustainability scientists or policy makers. What’s worse is that the biases introduced through several disciplines (and major journals) all go the same way, namely shedding light on coarse scales at the expense of finer scales.
Coarse-scale analyses are equivalent to aggregation of finer-scaled phenomena. Such aggregation happens in geographical space as well as in time, and also in thematic resolution. For example, ecological data may be scaled to large grid cells, or socioeconomic data may be scaled to the level of nations. Food production may be calculated for average years (ignoring the highs and lows through time), and households may be lumped into cities. Such aggregation is valid, and can be useful, or even important. But it’s also important to recognize that different patterns will be revealed by analyses at finer scales.
In the ecosystem services literature, this was recognized (for example) by Daw et al., who discuss the importance of disaggregating the beneficiaries of ecosystem services. This is because it’s possible that a lot of ecosystem services are generated by a system, but they all flow to the same beneficiary. Such a pattern is masked, however, by aggregated analyses. In economic analyses, disaggregation has been used, too, for example, when investigating income inequalities in a given country. In ecological data, some regions may look like they are rich in biodiversity when in fact, all of that diversity is clustered in a small area, such as a protected area (i.e. most of the region has low biodiversity). In terms of biophysical patterns, aggregated data tell us, for example, that Eastern Europe is a region with high yield gaps – but locally, this may or may not be the case.
Coarse-scale relationships such as the ones listed above are often particularly appealing to policy makers – and then an entire “aggregated bunch of things” is treated with one method, even though there are potential differences within. For example, it can be shown that high economic growth at the national level is related to national level food security. It follows then (at least at first glance) that any policy that boosts economic growth ought to have positive effects on food security. Or take land sparing: getting the highest biodiversity in a given region (can be achieved by creating a nature reserve and intensifying the land around it. At a coarse scale, at least in the first instance, this maximizes biodiversity.
The problem is that major journals and policy makers are a little too excited about such coarse-scale patterns. It’s not like they’re wrong, but they also don’t tell a very complete story. It pays to consider the effects of potential actions at multiple scales. For example, in the pursuit of economic growth, are certain stakeholders left behind? In generating wealth out of ecosystems, do all households in an area benefit equally? When protecting biodiversity, are small pockets of high diversity created or are entire landscapes managed to be ecologically self-sustaining?
So, by all means, we should look at coarse-scale patterns. Their generality is appealing and if we didn’t aggregate, we’d be entirely stuck. But coarse-scale patterns result from the aggregation of finer-scaled phenomena, and we should keep this in mind. In a way, it’s not only the “mean” outcome, but also its “distribution” that matters. Communicating this is difficult, but both leading journals and policy makers would be well advised to consider disaggregated analyses alongside aggregated analyses to make wise (editorial or policy) decisions.