The good, the bad, and the ugly of regenerative farming

Below are the key takeaways and introduction to our research into regenerative farming. Read the full report.

Our key takeaways

• Regenerative farming can refer to many different practices: one farm might rotate crops while limiting synthetic inputs and tillage whereas another might implement an agroforestry approach with resilient plants. As a first step, we reviewed multiple literature reviews on regenerative farming to identify the eleven most common practices. These are: agroforestry, reduced tillage, cover crops, rotational grazing, reduced synthetic inputs, crop rotation, composting and manure, stubble retention, polyculture, perennials, and mixed farming.
• In a second step, we separately reviewed the evidence of these 11 most common practices on five outcomes: biodiversity, carbon sequestration, income, soil health, and agricultural runoff. The strength of the evidence varies substantially on average across practices and across outcomes within each practice (see Figure 1). At one end of the spectrum, there is strong evidence supporting the benefits of certain practices on multiple outcomes. Agroforestry, for example, enhances biodiversity and carbon sequestration, and reduces runoff. At the other end, other methods have limited empirical support overall (e.g., mixed farming) or are beneficial for specific outcomes only (e.g., manure usage primarily boosting soil organic carbon). Guided by the high variance of the evidence, this report segments the practices into those with strong evidence (5 practices), medium-quality evidence (2 practices), and weak evidence (4 practices).
• The impact of most regenerative agriculture techniques on income varies substantially across studies and contexts. Cover crops might be beneficial in humid climates but draw too much water from the ground in arid climates. Similarly, using manure might increase farm income if local manure supply is high and the nutrient profile of crops demands it but is unprofitable otherwise. The income effects will determine whether it is in farmers’ interest to implement regenerative farming. As such, when analyzing the potential of different interventions or funding opportunities, it is crucial to determine whether the specific conditions of that intervention are likely to increase or reduce farmers’ profits.

An introduction

Regenerative farming describes practices that aim to improve farmland on various dimensions, including regenerating top soil, improving biodiversity, and making agriculture more sustainable so that land can be cultivated over the long run. Because regenerative farming is often described by its goals, rather than its specific components, there is no generally agreed-upon set of practices that make up regenerative farming. This lack of a clear definition of its components poses a difficulty for research that has been pointed out in various review studies (e.g., Newton et al. 2020, Schreefel et al. 2015): if practitioners refer to different techniques when implementing regenerative farming, researchers do not have a well-defined concept to study. As a result, existing research has instead focused on more clearly defined agricultural techniques, such as crop rotations or tillage, which might make up regenerative farming in certain instances. This research lens, however, limits the academic evidence on the effects of regenerative farming as a whole. To nonetheless incorporate all available evidence on regenerative farming, this report reviews the various practices that might constitute regenerative farming and estimates their effects on various outcomes. Those interested in the effect of a particular conception of regenerative farming—e.g., practices #1, #5, and #10, but not others—may sum the individual effects of the different practices listed in this report to approximate the overall effect of their combination.

To gather and combine the evidence on regenerative farming, we began by reviewing the academic literature to identify its most common elements: Khangura et al. (2023) and Schreefel et al. (2015) both combine data from previous regenerative farming studies and distill common objectives and activities. We combine their results into different categories of regenerative farming techniques depending on their primary focus on the farm¹:

Soil focus:

• Reduced tillage
• Stubble retention (retaining the straw and crown of plants left after harvesting)
• Cover crops (planting crops to cover the soil in between growing seasons)

Grazing focus:

• Rotational grazing
• Mixed farming (integrating crops and livestock on the same land)

Crop focus:

• Polyculture (planting different crops alongside each other)
• Crop rotation (growing different crops in different growing seasons)
• Use of perennials

Inputs focus:

• Minimize synthetic inputs
• Composting and manure

Miscellaneous

• Agroforestry

We then conducted a review of the academic literature on the effects of each practice on five outcomes: biodiversity, carbon sequestration, income, soil health, and agricultural runoff. For this process, we classify the evidence quality into strong, medium-quality, and weak evidence. Strong evidence generally consists of a range of academic studies in different contexts, meta-analyses or systematic reviews that suggest both high internal and external validity of the findings (i.e., results are likely to both be an accurate representation of the effects that the study aims to estimate, and are likely to generalize to different contexts). Medium-quality describes an evidence base characterized by multiple individual studies that are overall limited in scope or only focus on certain contexts (e.g., the research on the biodiversity effects of cover crops consisting of multiple studies but being mostly limited to vineyards and arthropods). Weak evidence refers to a strand of literature that has few if any credible studies and thus does not give us much certainty about the effectiveness of a given practice (e.g., the effects of stubble retention on biodiversity being mostly limited to one study of farmland birds in the UK).

Figure 1 (see full report) lists the result of this evidence review in a compact way. For each practice and outcome, we list the direction of an effect (positive/neutral/negative) as well as the quality of the evidence supporting it (strong/medium-quality/weak). It is important to note, however, that we’ve not included indicators of effect sizes (i.e., how positive or negative an effect is). While a practice might have very certain positive effects, those effects might nonetheless be small in magnitude. For philanthropists wishing to compare interventions based on their cost-effectiveness to achieve a certain goal it would be crucial to understand the size of effects. To put it concretely, a philanthropist interested in the most effective ways to increase carbon sequestration should aim to compare the effect size both within various regenerative farming practices as well as across regenerative farming and promising non-regenerative farming interventions. While it is not in the scope of this report to compare the effect sizes of different regenerative farming practices, it would be an interesting question for further study. Depending on a philanthropist’s particular objectives, a prioritization might lead to very different results.

The remainder of the report discusses the effects of each practice on each outcome in turn. Firstly, we discuss practices with strong evidence: agroforestry, reduced tillage, cover crops, rotational grazing, and reduced synthetic inputs. We then explore practices with medium-quality evidence (crop rotation and composting/manure), before assessing the effects of practices with weak evidence: stubble retention, polyculture, perennials, and mixed farming. To categorize a practice into an evidence quality category, we assess its overall evidence base across outcomes. As such, a practice with strong evidence, for example, might have very strong evidence for one outcome but weaker evidence for another. This aggregate categorization is designed to provide an overall view on the evidence base of different activities.

Read the full report.

Notes

1. Though this is not a definite categorization. Various practices might just as well fit into other categories, especially if their purpose on a specific farm is different. ↩