An application of causal ML to identify the heterogeneous effects of AES

class: center, middle, inverse, title-slide

# An application of causal ML to identify the heterogeneous effects of AES
## A case study in the EU
### Christian Stetter
### Technical University Munich <br /> 2022/03/07 <br />
### <br /> Slides available at: rpubs.com/cstetter/aecp

---

# My research agenda: <br/> Agricultural production and environmental change

.pull-left[
<img src="data:image/png;base64,#concept.png" width="500px" />
]
--

.pull-right[
### Selected previous & current projects
.font70[
- Stetter C, Sauer J (2022). Greenhouse Gas Emissions and Eco-Performance at Farm Level: A Parametric Approach.Environmental and Resource Economics.

- Stetter C, Wimmer S, Sauer J (In press). Production Intensity and Emission Efficiency – A Latent-Class Meta-Frontier Approach. Journal of Agricultural and Resource Economics.

- Stetter C, Sauer J (2021). Tackling Climate Change: Agroforestry Adoption in the Face
of Regional Weather Extremes (working paper).

- Stetter C, Sauer, J (2021). Exploring the heterogeneous effects of weather on productivity
using generalized random forests (working paper).

- Stetter C, Mennig P, Sauer J (forthcoming). **Using Machine Learning to Identify Heterogeneous Impacts of Agri-Environment Schemes in the EU: A Case Study.** European Review of Agricultural Economics.
]]

---
class: hide-count, middle center
count: false
# Today's talk: <br/> Causal machine learning to identify the heterogeneous effects of agri-environmental schemes in the EU

---
# Farming and the environment in the European Union
- Like other regions in the world, farming in the EU is faced with several environmental challenges:

- Loss of insects 
       - Soil erosion 
       - Water pollution 
       - GHG emissions
       
<br/>

- There has been lively debate on how the Common Agricultural Policy of the EU can tackle these challenges.

<br/>

- Consensus about key objectives of the CAP in terms of the environment.

- Environmental care
    - Climate change action 
    - Preservation of landscapes
    - Biodiversity

---
# What are agri-environmental schemes

- As a means to reduce the pressure of agriculture on the environment, the EU introduced agri-environmental schemes

- Not part of direct payments system

- Voluntary

- **How they work (action-based):**

- Farmers commit themselves to adopt environmentally-friendly farming techniques that go beyond legal obligations.
      
      - In return, they receive payments that are supposed to provide compensation for additional costs and income foregone (resulting from applying those environmentally friendly farming practices)

- **Examples:**

- Environmentally favorable extensification of farming 
    - Five-part crop rotation

.right[**...but are they actually effective?**]

---
# Introduction: What does the literature say?

**Econometric farm-level studies in Europe:**

.pull-left[

.font90[
- **Average** treatment effects of AES on, e.g.:

- Fertilizer and pesticide use [0/-] <a name=cite-Uehleke2022></a>[[1](https://doi.org/10.1016/j.landusepol.2021.105950)]
    <a name=cite-Pufahl2009></a>[[2](https://doi.org/10.1093/erae/jbp001)]
    <a name=cite-Arata2016></a>[[3](https://doi.org/10.3368/le.92.1.167)]
    
    - Crop diversity [0/+] <a name=cite-Bertoni2020></a>[[4](https://doi.org/10.1016/j.foodpol.2019.101790)]
    
    - Grassland (share/diversity)  [+] [[1](https://doi.org/10.1016/j.landusepol.2021.105950)]  [[3](https://doi.org/10.3368/le.92.1.167)]
    
    - Biodiversity conservation area [+] <a name=cite-Wuepper2021></a>[[5](https://doi.org/10.1111/ajae.12284)]
]

<br/>
- Usual estimation approaches:

- Matching
    
    - Difference in Difference (DID)
    
    - DID Matching
]

.pull-right[
<br/>
<br/>
...**but**:
- *AES might have different effects on different farms*

- *These methods are not built for assessing heterogeneous effects*
]

---
# Research Objectives

### 1. Find a suitable approach to estimate heterogeneous effects of AES participation and demonstrate its usefulness.

<br/>

### 2. Assess the impact heterogeneity of AES within the climate, biodiversity, soil and water health domains.

<br/>

### 3. .gray[Use contextual knowledge to improve targeting for an increased effectiveness of AES.]
.gray[
- See [Appendix Targeting](#app_interpretation)]

---
# Theoretical Framework: Production possibilities 
.left-column[
Production possibility curve: 
- All possible combinations of two outputs that could potentially be produced at given input bundle

Iso-revenue line:
- All possible combinations of two products which would yield the same total revenue

Farm produces at point A0 where PPC and IR are tangent
]

.right-column[
<img src="data:image/png;base64,#img/Folie1.PNG" width="95%" />
]

---
count: false
# Theoretical Framework: Production possibilities
.left-column[
<br/>
<br/>
<br/>
- Action-based AES limit production possibilities

- Downward shift of PPC under AES participation

]

.right-column[
<img src="data:image/png;base64,#img/Folie2.PNG" width="95%" />
]

---
count: false
# Theoretical Framework: Production possibilities
.left-column[
<br/>
<br/>
<br/>
- Farm produces at point A1 where PPC and IR are tangent

- Two alternative points of production

- Two potential outcomes
]

.right-column[
<img src="data:image/png;base64,#img/Folie3.PNG" width="95%" />
]

---
count: false
# Theoretical Framework: Potential outcomes
.left-column[
<br/>
<br/>
<br/>
Treatment effect:
- Expected difference between two potential outcomes

- Fundamental problem of causal inference: only one outcome observable

]

.right-column[
<img src="data:image/png;base64,#img/Folie4.PNG" width="95%" />
]

---
# Theoretical Framework: Conditional average treatment effect
![](data:image/png;base64,#presentation_files/figure-html/unnamed-chunk-6-1.png)

--
.pull-left[
- Production possibilities depend on farm-specific contextual variables `$X$`

- Effect of interest: `$$\tau = E[Q_1 - Q_0| X=x ]$$`
]

.pull-right[
**Contextual domains (defining `$X$`):**
- Resource bundle and input intensities
- Output bundle
- Farm and farmer characteristics
- Biophysical environment
- Institutional and market environment
]

---
# Estimation strategy: Generalized random forest
.pull-left[
- Uses the basic ideas of regression trees and **random forests**:

- Recursive partitioning, subsampling and random split selection of input variables

- Subsample is recursively divided into binary nodes such that the prediction error is minimized

]

.pull-right[
**Generalized random forest <a name=cite-Athey2019f></a>[[6](https://doi.org/10.1214/18-AOS1709)]**
- **Adjusted splitting rule**: 
      
    - Instead of minimizing the prediction error
    - Maximize homogeneity within each subgroup and maximize treatment effect across between subgroups

- Allows to assess statistical uncertainty

- Allows for (higher-order) interactions and nonlinearities

- De facto no limit on the number of predictors

<br/>
.right[[More Details](#app_grf)]]

---
# Data
.pull-left[
**Focus region: Bavaria**

- Farm accountancy data

- Sample of 2758 farms for 2014

**Environmental indicators (Y)**
<div id="siiqociizf" style="overflow-x:auto;overflow-y:auto;width:auto;height:auto;">
<style>html {
  font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, Oxygen, Ubuntu, Cantarell, 'Helvetica Neue', 'Fira Sans', 'Droid Sans', Arial, sans-serif;
}

#siiqociizf .gt_table {
  display: table;
  border-collapse: collapse;
  margin-left: auto;
  margin-right: auto;
  color: #333333;
  font-size: 16px;
  font-weight: normal;
  font-style: normal;
  background-color: #FFFFFF;
  width: auto;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #A8A8A8;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #A8A8A8;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
}

#siiqociizf .gt_heading {
  background-color: #FFFFFF;
  text-align: center;
  border-bottom-color: #FFFFFF;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
}

#siiqociizf .gt_title {
  color: #333333;
  font-size: 125%;
  font-weight: initial;
  padding-top: 4px;
  padding-bottom: 4px;
  padding-left: 5px;
  padding-right: 5px;
  border-bottom-color: #FFFFFF;
  border-bottom-width: 0;
}

#siiqociizf .gt_subtitle {
  color: #333333;
  font-size: 85%;
  font-weight: initial;
  padding-top: 0;
  padding-bottom: 6px;
  padding-left: 5px;
  padding-right: 5px;
  border-top-color: #FFFFFF;
  border-top-width: 0;
}

#siiqociizf .gt_bottom_border {
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#siiqociizf .gt_col_headings {
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
}

#siiqociizf .gt_col_heading {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: normal;
  text-transform: inherit;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: bottom;
  padding-top: 5px;
  padding-bottom: 6px;
  padding-left: 5px;
  padding-right: 5px;
  overflow-x: hidden;
}

#siiqociizf .gt_column_spanner_outer {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: normal;
  text-transform: inherit;
  padding-top: 0;
  padding-bottom: 0;
  padding-left: 4px;
  padding-right: 4px;
}

#siiqociizf .gt_column_spanner_outer:first-child {
  padding-left: 0;
}

#siiqociizf .gt_column_spanner_outer:last-child {
  padding-right: 0;
}

#siiqociizf .gt_column_spanner {
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  vertical-align: bottom;
  padding-top: 5px;
  padding-bottom: 5px;
  overflow-x: hidden;
  display: inline-block;
  width: 100%;
}

#siiqociizf .gt_group_heading {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: middle;
}

#siiqociizf .gt_empty_group_heading {
  padding: 0.5px;
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  vertical-align: middle;
}

#siiqociizf .gt_from_md > :first-child {
  margin-top: 0;
}

#siiqociizf .gt_from_md > :last-child {
  margin-bottom: 0;
}

#siiqociizf .gt_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  margin: 10px;
  border-top-style: solid;
  border-top-width: 1px;
  border-top-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: middle;
  overflow-x: hidden;
}

#siiqociizf .gt_stub {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-right-style: solid;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  padding-left: 5px;
  padding-right: 5px;
}

#siiqociizf .gt_stub_row_group {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-right-style: solid;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  padding-left: 5px;
  padding-right: 5px;
  vertical-align: top;
}

#siiqociizf .gt_row_group_first td {
  border-top-width: 2px;
}

#siiqociizf .gt_summary_row {
  color: #333333;
  background-color: #FFFFFF;
  text-transform: inherit;
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
}

#siiqociizf .gt_first_summary_row {
  border-top-style: solid;
  border-top-color: #D3D3D3;
}

#siiqociizf .gt_first_summary_row.thick {
  border-top-width: 2px;
}

#siiqociizf .gt_last_summary_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#siiqociizf .gt_grand_summary_row {
  color: #333333;
  background-color: #FFFFFF;
  text-transform: inherit;
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
}

#siiqociizf .gt_first_grand_summary_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  border-top-style: double;
  border-top-width: 6px;
  border-top-color: #D3D3D3;
}

#siiqociizf .gt_striped {
  background-color: rgba(128, 128, 128, 0.05);
}

#siiqociizf .gt_table_body {
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#siiqociizf .gt_footnotes {
  color: #333333;
  background-color: #FFFFFF;
  border-bottom-style: none;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
}

#siiqociizf .gt_footnote {
  margin: 0px;
  font-size: 90%;
  padding-left: 4px;
  padding-right: 4px;
  padding-left: 5px;
  padding-right: 5px;
}

#siiqociizf .gt_sourcenotes {
  color: #333333;
  background-color: #FFFFFF;
  border-bottom-style: none;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
}

#siiqociizf .gt_sourcenote {
  font-size: 90%;
  padding-top: 4px;
  padding-bottom: 4px;
  padding-left: 5px;
  padding-right: 5px;
}

#siiqociizf .gt_left {
  text-align: left;
}

#siiqociizf .gt_center {
  text-align: center;
}

#siiqociizf .gt_right {
  text-align: right;
  font-variant-numeric: tabular-nums;
}

#siiqociizf .gt_font_normal {
  font-weight: normal;
}

#siiqociizf .gt_font_bold {
  font-weight: bold;
}

#siiqociizf .gt_font_italic {
  font-style: italic;
}

#siiqociizf .gt_super {
  font-size: 65%;
}

#siiqociizf .gt_footnote_marks {
  font-style: italic;
  font-weight: normal;
  font-size: 75%;
  vertical-align: 0.4em;
}

#siiqociizf .gt_asterisk {
  font-size: 100%;
  vertical-align: 0;
}

#siiqociizf .gt_slash_mark {
  font-size: 0.7em;
  line-height: 0.7em;
  vertical-align: 0.15em;
}

#siiqociizf .gt_fraction_numerator {
  font-size: 0.6em;
  line-height: 0.6em;
  vertical-align: 0.45em;
}

#siiqociizf .gt_fraction_denominator {
  font-size: 0.6em;
  line-height: 0.6em;
  vertical-align: -0.05em;
}
</style>
<table class="gt_table" style="table-layout: fixed;; width: 0px">
  <colgroup>
    <col style="width:110px;"/>
    <col style="width:170px;"/>
    <col style="width:60px;"/>
    <col style="width:60px;"/>
    <col style="width:60px;"/>
  </colgroup>
  
  <thead class="gt_col_headings">
    <tr>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 16px;">Domain</th>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 16px;">Indicator</th>
      <th class="gt_col_heading gt_columns_bottom_border gt_right" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 16px;">Teated</th>
      <th class="gt_col_heading gt_columns_bottom_border gt_right" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 16px;">Un-treated</th>
      <th class="gt_col_heading gt_columns_bottom_border gt_right" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 16px;">Entire sample</th>
    </tr>
  </thead>
  <tbody class="gt_table_body">
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Soil/water</div></td>
<td class="gt_row gt_left"><div><span style='font-weight:bold;font-size:14px;color:#454545'>Fertilizer intensity</div>
<div><span style='font-size:14px;color:#454545'>(Euro/ha)</div></td>
<td class="gt_row gt_right">187</td>
<td class="gt_row gt_right">205</td>
<td class="gt_row gt_right">194</td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Soil/water</div></td>
<td class="gt_row gt_left"><div><span style='font-weight:bold;font-size:14px;color:#454545'>Pesticide intensity</div>
<div><span style='font-size:14px;color:#454545'>(Euro/ha)</div></td>
<td class="gt_row gt_right">120</td>
<td class="gt_row gt_right">121</td>
<td class="gt_row gt_right">120</td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Biodiversity</div></td>
<td class="gt_row gt_left"><div><span style='font-weight:bold;font-size:14px;color:#454545'>Gini-Simpson index</div>
<div><span style='font-size:14px;color:#454545'>(0-100)</div></td>
<td class="gt_row gt_right">67</td>
<td class="gt_row gt_right">64</td>
<td class="gt_row gt_right">66</td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Climate</div></td>
<td class="gt_row gt_left"><div><span style='font-weight:bold;font-size:14px;color:#454545'>GHG emissions</div>
<div><span style='font-size:14px;color:#454545'>(t CO2eq )</div></td>
<td class="gt_row gt_right">469</td>
<td class="gt_row gt_right">411</td>
<td class="gt_row gt_right">445</td></tr>
  </tbody>
  
  
</table>
</div>
]

.pull-right[
**Treatment (W)**

- Binary participation indicator (Treated: 1677 Untreated: 1081)

**Contextual variables (X, ca. 130)**

<div id="kqyauwxmaq" style="overflow-x:auto;overflow-y:auto;width:auto;height:auto;">
<style>html {
  font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, Oxygen, Ubuntu, Cantarell, 'Helvetica Neue', 'Fira Sans', 'Droid Sans', Arial, sans-serif;
}

#kqyauwxmaq .gt_table {
  display: table;
  border-collapse: collapse;
  margin-left: auto;
  margin-right: auto;
  color: #333333;
  font-size: 16px;
  font-weight: normal;
  font-style: normal;
  background-color: #FFFFFF;
  width: auto;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #A8A8A8;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #A8A8A8;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
}

#kqyauwxmaq .gt_heading {
  background-color: #FFFFFF;
  text-align: center;
  border-bottom-color: #FFFFFF;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
}

#kqyauwxmaq .gt_title {
  color: #333333;
  font-size: 125%;
  font-weight: initial;
  padding-top: 4px;
  padding-bottom: 4px;
  padding-left: 5px;
  padding-right: 5px;
  border-bottom-color: #FFFFFF;
  border-bottom-width: 0;
}

#kqyauwxmaq .gt_subtitle {
  color: #333333;
  font-size: 85%;
  font-weight: initial;
  padding-top: 0;
  padding-bottom: 6px;
  padding-left: 5px;
  padding-right: 5px;
  border-top-color: #FFFFFF;
  border-top-width: 0;
}

#kqyauwxmaq .gt_bottom_border {
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#kqyauwxmaq .gt_col_headings {
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
}

#kqyauwxmaq .gt_col_heading {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: normal;
  text-transform: inherit;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: bottom;
  padding-top: 5px;
  padding-bottom: 6px;
  padding-left: 5px;
  padding-right: 5px;
  overflow-x: hidden;
}

#kqyauwxmaq .gt_column_spanner_outer {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: normal;
  text-transform: inherit;
  padding-top: 0;
  padding-bottom: 0;
  padding-left: 4px;
  padding-right: 4px;
}

#kqyauwxmaq .gt_column_spanner_outer:first-child {
  padding-left: 0;
}

#kqyauwxmaq .gt_column_spanner_outer:last-child {
  padding-right: 0;
}

#kqyauwxmaq .gt_column_spanner {
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  vertical-align: bottom;
  padding-top: 5px;
  padding-bottom: 5px;
  overflow-x: hidden;
  display: inline-block;
  width: 100%;
}

#kqyauwxmaq .gt_group_heading {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: middle;
}

#kqyauwxmaq .gt_empty_group_heading {
  padding: 0.5px;
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  vertical-align: middle;
}

#kqyauwxmaq .gt_from_md > :first-child {
  margin-top: 0;
}

#kqyauwxmaq .gt_from_md > :last-child {
  margin-bottom: 0;
}

#kqyauwxmaq .gt_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  margin: 10px;
  border-top-style: solid;
  border-top-width: 1px;
  border-top-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: middle;
  overflow-x: hidden;
}

#kqyauwxmaq .gt_stub {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-right-style: solid;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  padding-left: 5px;
  padding-right: 5px;
}

#kqyauwxmaq .gt_stub_row_group {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-right-style: solid;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  padding-left: 5px;
  padding-right: 5px;
  vertical-align: top;
}

#kqyauwxmaq .gt_row_group_first td {
  border-top-width: 2px;
}

#kqyauwxmaq .gt_summary_row {
  color: #333333;
  background-color: #FFFFFF;
  text-transform: inherit;
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
}

#kqyauwxmaq .gt_first_summary_row {
  border-top-style: solid;
  border-top-color: #D3D3D3;
}

#kqyauwxmaq .gt_first_summary_row.thick {
  border-top-width: 2px;
}

#kqyauwxmaq .gt_last_summary_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#kqyauwxmaq .gt_grand_summary_row {
  color: #333333;
  background-color: #FFFFFF;
  text-transform: inherit;
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
}

#kqyauwxmaq .gt_first_grand_summary_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  border-top-style: double;
  border-top-width: 6px;
  border-top-color: #D3D3D3;
}

#kqyauwxmaq .gt_striped {
  background-color: rgba(128, 128, 128, 0.05);
}

#kqyauwxmaq .gt_table_body {
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#kqyauwxmaq .gt_footnotes {
  color: #333333;
  background-color: #FFFFFF;
  border-bottom-style: none;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
}

#kqyauwxmaq .gt_footnote {
  margin: 0px;
  font-size: 90%;
  padding-left: 4px;
  padding-right: 4px;
  padding-left: 5px;
  padding-right: 5px;
}

#kqyauwxmaq .gt_sourcenotes {
  color: #333333;
  background-color: #FFFFFF;
  border-bottom-style: none;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
}

#kqyauwxmaq .gt_sourcenote {
  font-size: 90%;
  padding-top: 4px;
  padding-bottom: 4px;
  padding-left: 5px;
  padding-right: 5px;
}

#kqyauwxmaq .gt_left {
  text-align: left;
}

#kqyauwxmaq .gt_center {
  text-align: center;
}

#kqyauwxmaq .gt_right {
  text-align: right;
  font-variant-numeric: tabular-nums;
}

#kqyauwxmaq .gt_font_normal {
  font-weight: normal;
}

#kqyauwxmaq .gt_font_bold {
  font-weight: bold;
}

#kqyauwxmaq .gt_font_italic {
  font-style: italic;
}

#kqyauwxmaq .gt_super {
  font-size: 65%;
}

#kqyauwxmaq .gt_footnote_marks {
  font-style: italic;
  font-weight: normal;
  font-size: 75%;
  vertical-align: 0.4em;
}

#kqyauwxmaq .gt_asterisk {
  font-size: 100%;
  vertical-align: 0;
}

#kqyauwxmaq .gt_slash_mark {
  font-size: 0.7em;
  line-height: 0.7em;
  vertical-align: 0.15em;
}

#kqyauwxmaq .gt_fraction_numerator {
  font-size: 0.6em;
  line-height: 0.6em;
  vertical-align: 0.45em;
}

#kqyauwxmaq .gt_fraction_denominator {
  font-size: 0.6em;
  line-height: 0.6em;
  vertical-align: -0.05em;
}
</style>
<table class="gt_table" style="table-layout: fixed;; width: 0px">
  <colgroup>
    <col style="width:150px;"/>
    <col style="width:300px;"/>
  </colgroup>
  
  <thead class="gt_col_headings">
    <tr>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 16px;">Domain</th>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 16px;">Description</th>
    </tr>
  </thead>
  <tbody class="gt_table_body">
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Resource bundle and input intensities</div></td>
<td class="gt_row gt_left"><div><span style='font-size:14px;color:#454545'>Land use, labor, materials & capital, 
                         cultivation plan, livestock count and composition</div></td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Output bundle</div></td>
<td class="gt_row gt_left"><div><span style='font-size:14px;color:#454545'>Crop and livestock outputs</div></td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Farm and farmer characteristics</div></td>
<td class="gt_row gt_left"><div><span style='font-size:14px;color:#454545'>Farm type, whole farm value added, farmer's age, yields etc.</div></td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Biophysical environment</div></td>
<td class="gt_row gt_left"><div><span style='font-size:14px;color:#454545'>Soil quality, altitude etc.</div></td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px;color:black'>Institutional and market environment</div></td>
<td class="gt_row gt_left"><div><span style='font-size:14px;color:#454545'>Administrative district, local unemployment rate, 
                         local land rental price etc</div></td></tr>
  </tbody>
  
  
</table>
</div>
]

---
# Results: Heterogeneous impact predictions
<div id="htmlwidget-0646be29c13e692fd68e" style="width:100%;height:70%;" class="plotly html-widget"></div>
<script type="application/json" data-for="htmlwidget-0646be29c13e692fd68e">{"x":{"data":[{"orientation":"v","width":[3.41767523621459,3.41767523621459,3.41767523621459,3.41767523621458,3.41767523621459,3.41767523621459,3.41767523621459,3.41767523621459,3.41767523621459,3.41767523621459,3.41767523621458,3.41767523621459,3.41767523621458,3.41767523621459,3.41767523621458,3.41767523621459,3.41767523621459,3.41767523621458,3.41767523621459,3.41767523621459,3.41767523621458,3.41767523621459,3.41767523621459,3.41767523621459,3.41767523621459,3.41767523621458,3.41767523621459,3.41767523621459,3.41767523621459,3.41767523621458],"base":[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"x":[-37.5944275983605,-34.1767523621459,-30.7590771259313,-27.3414018897167,-23.9237266535021,-20.5060514172875,-17.0883761810729,-13.6707009448584,-10.2530257086438,-6.83535047242918,-3.41767523621459,0,3.41767523621459,6.83535047242918,10.2530257086438,13.6707009448584,17.0883761810729,20.5060514172875,23.9237266535021,27.3414018897167,30.7590771259313,34.1767523621459,37.5944275983605,41.0121028345751,44.4297780707897,47.8474533070042,51.2651285432188,54.6828037794334,58.100479015648,61.5181542518626],"y":[1,1,1,1,1,6,7,22,52,154,311,553,643,456,265,103,68,29,13,18,7,5,4,4,3,2,1,1,1,1],"text":["count:   1<br />value: -3.759443e+01<br />name: GHG emissions (t)","count:   1<br />value: -3.417675e+01<br />name: GHG emissions (t)","count:   1<br />value: -3.075908e+01<br />name: GHG emissions (t)","count:   1<br />value: -2.734140e+01<br />name: GHG emissions (t)","count:   1<br />value: -2.392373e+01<br />name: GHG emissions (t)","count:   6<br />value: -2.050605e+01<br />name: GHG emissions (t)","count:   7<br />value: -1.708838e+01<br />name: GHG emissions (t)","count:  22<br />value: -1.367070e+01<br />name: GHG emissions (t)","count:  52<br />value: -1.025303e+01<br />name: GHG emissions (t)","count: 154<br />value: -6.835350e+00<br />name: GHG emissions (t)","count: 311<br />value: -3.417675e+00<br />name: GHG emissions (t)","count: 553<br />value:  0.000000e+00<br />name: GHG emissions (t)","count: 643<br />value:  3.417675e+00<br />name: GHG emissions (t)","count: 456<br />value:  6.835350e+00<br />name: GHG emissions (t)","count: 265<br />value:  1.025303e+01<br />name: GHG emissions (t)","count: 103<br />value:  1.367070e+01<br />name: GHG emissions (t)","count:  68<br />value:  1.708838e+01<br />name: GHG emissions (t)","count:  29<br />value:  2.050605e+01<br />name: GHG emissions (t)","count:  13<br />value:  2.392373e+01<br />name: GHG emissions (t)","count:  18<br />value:  2.734140e+01<br />name: GHG emissions (t)","count:   7<br />value:  3.075908e+01<br />name: GHG emissions (t)","count:   5<br />value:  3.417675e+01<br />name: GHG emissions (t)","count:   4<br />value:  3.759443e+01<br />name: GHG emissions (t)","count:   4<br />value:  4.101210e+01<br />name: GHG emissions (t)","count:   3<br />value:  4.442978e+01<br />name: GHG emissions (t)","count:   2<br />value:  4.784745e+01<br />name: GHG emissions (t)","count:   1<br />value:  5.126513e+01<br />name: GHG emissions (t)","count:   1<br />value:  5.468280e+01<br />name: GHG emissions (t)","count:   1<br />value:  5.810048e+01<br />name: GHG emissions (t)","count:   1<br />value:  6.151815e+01<br />name: GHG emissions (t)"],"type":"bar","textposition":"none","marker":{"autocolorscale":false,"color":"rgba(248,118,109,1)","line":{"width":1.88976377952756,"color":"transparent"}},"name":"GHG emissions (t)","legendgroup":"GHG emissions (t)","showlegend":true,"xaxis":"x2","yaxis":"y2","hoverinfo":"text","frame":null},{"orientation":"v","width":[1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656,1.7031605679656],"base":[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"x":[-30.6568902233809,-28.9537296554153,-27.2505690874496,-25.547408519484,-23.8442479515184,-22.1410873835528,-20.4379268155872,-18.7347662476216,-17.031605679656,-15.3284451116904,-13.6252845437248,-11.9221239757592,-10.2189634077936,-8.51580283982801,-6.81264227186241,-5.10948170389681,-3.4063211359312,-1.7031605679656,1.77635683940025e-15,1.7031605679656,3.40632113593121,5.10948170389681,6.81264227186242,8.51580283982802,10.2189634077936,11.9221239757592,13.6252845437248,15.3284451116904,17.031605679656,18.7347662476216],"y":[2,3,7,8,20,31,37,82,121,181,258,329,338,301,281,211,184,131,81,59,28,16,13,3,3,1,3,0,0,2],"text":["count:   2<br />value: -3.065689e+01<br />name: Fertilizer intensity (EUR/ha)","count:   3<br />value: -2.895373e+01<br />name: Fertilizer intensity (EUR/ha)","count:   7<br />value: -2.725057e+01<br />name: Fertilizer intensity (EUR/ha)","count:   8<br />value: -2.554741e+01<br />name: Fertilizer intensity (EUR/ha)","count:  20<br />value: -2.384425e+01<br />name: Fertilizer intensity (EUR/ha)","count:  31<br />value: -2.214109e+01<br />name: Fertilizer intensity (EUR/ha)","count:  37<br />value: -2.043793e+01<br />name: Fertilizer intensity (EUR/ha)","count:  82<br />value: -1.873477e+01<br />name: Fertilizer intensity (EUR/ha)","count: 121<br />value: -1.703161e+01<br />name: Fertilizer intensity (EUR/ha)","count: 181<br />value: -1.532845e+01<br />name: Fertilizer intensity (EUR/ha)","count: 258<br />value: -1.362528e+01<br />name: Fertilizer intensity (EUR/ha)","count: 329<br />value: -1.192212e+01<br />name: Fertilizer intensity (EUR/ha)","count: 338<br />value: -1.021896e+01<br />name: Fertilizer intensity (EUR/ha)","count: 301<br />value: -8.515803e+00<br />name: Fertilizer intensity (EUR/ha)","count: 281<br />value: -6.812642e+00<br />name: Fertilizer intensity (EUR/ha)","count: 211<br />value: -5.109482e+00<br />name: Fertilizer intensity (EUR/ha)","count: 184<br />value: -3.406321e+00<br />name: Fertilizer intensity (EUR/ha)","count: 131<br />value: -1.703161e+00<br />name: Fertilizer intensity (EUR/ha)","count:  81<br />value:  1.776357e-15<br />name: Fertilizer intensity (EUR/ha)","count:  59<br />value:  1.703161e+00<br />name: Fertilizer intensity (EUR/ha)","count:  28<br />value:  3.406321e+00<br />name: Fertilizer intensity (EUR/ha)","count:  16<br />value:  5.109482e+00<br />name: Fertilizer intensity (EUR/ha)","count:  13<br />value:  6.812642e+00<br />name: Fertilizer intensity (EUR/ha)","count:   3<br />value:  8.515803e+00<br />name: Fertilizer intensity (EUR/ha)","count:   3<br />value:  1.021896e+01<br />name: Fertilizer intensity (EUR/ha)","count:   1<br />value:  1.192212e+01<br />name: Fertilizer intensity (EUR/ha)","count:   3<br />value:  1.362528e+01<br />name: Fertilizer intensity (EUR/ha)","count:   0<br />value:  1.532845e+01<br />name: Fertilizer intensity (EUR/ha)","count:   0<br />value:  1.703161e+01<br />name: Fertilizer intensity (EUR/ha)","count:   2<br />value:  1.873477e+01<br />name: Fertilizer intensity (EUR/ha)"],"type":"bar","textposition":"none","marker":{"autocolorscale":false,"color":"rgba(124,174,0,1)","line":{"width":1.88976377952756,"color":"transparent"}},"name":"Fertilizer intensity (EUR/ha)","legendgroup":"Fertilizer intensity (EUR/ha)","showlegend":true,"xaxis":"x","yaxis":"y","hoverinfo":"text","frame":null},{"orientation":"v","width":[2.75393613252448,2.75393613252447,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252447,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252448,2.75393613252447,2.75393613252448,2.75393613252448,2.75393613252447,2.75393613252447,2.75393613252447,2.75393613252447,2.75393613252449,2.75393613252447,2.75393613252447,2.75393613252449,2.75393613252447],"base":[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"x":[-22.0314890601958,-19.2775529276714,-16.5236167951469,-13.7696806626224,-11.0157445300979,-8.26180839757344,-5.50787226504896,-2.75393613252448,0,2.75393613252448,5.50787226504896,8.26180839757344,11.0157445300979,13.7696806626224,16.5236167951469,19.2775529276713,22.0314890601958,24.7854251927203,27.5393613252448,30.2932974577693,33.0472335902937,35.8011697228182,38.5551058553427,41.3090419878672,44.0629781203917,46.8169142529161,49.5708503854406,52.3247865179651,55.0787226504896,57.832658783014],"y":[2,6,24,50,158,221,297,516,846,346,97,59,32,22,16,15,9,6,4,2,0,0,0,2,1,0,0,1,1,1],"text":["count:   2<br />value: -2.203149e+01<br />name: Pesticide intensity (EUR/ha)","count:   6<br />value: -1.927755e+01<br />name: Pesticide intensity (EUR/ha)","count:  24<br />value: -1.652362e+01<br />name: Pesticide intensity (EUR/ha)","count:  50<br />value: -1.376968e+01<br />name: Pesticide intensity (EUR/ha)","count: 158<br />value: -1.101574e+01<br />name: Pesticide intensity (EUR/ha)","count: 221<br />value: -8.261808e+00<br />name: Pesticide intensity (EUR/ha)","count: 297<br />value: -5.507872e+00<br />name: Pesticide intensity (EUR/ha)","count: 516<br />value: -2.753936e+00<br />name: Pesticide intensity (EUR/ha)","count: 846<br />value:  0.000000e+00<br />name: Pesticide intensity (EUR/ha)","count: 346<br />value:  2.753936e+00<br />name: Pesticide intensity (EUR/ha)","count:  97<br />value:  5.507872e+00<br />name: Pesticide intensity (EUR/ha)","count:  59<br />value:  8.261808e+00<br />name: Pesticide intensity (EUR/ha)","count:  32<br />value:  1.101574e+01<br />name: Pesticide intensity (EUR/ha)","count:  22<br />value:  1.376968e+01<br />name: Pesticide intensity (EUR/ha)","count:  16<br />value:  1.652362e+01<br />name: Pesticide intensity (EUR/ha)","count:  15<br />value:  1.927755e+01<br />name: Pesticide intensity (EUR/ha)","count:   9<br />value:  2.203149e+01<br />name: Pesticide intensity (EUR/ha)","count:   6<br />value:  2.478543e+01<br />name: Pesticide intensity (EUR/ha)","count:   4<br />value:  2.753936e+01<br />name: Pesticide intensity (EUR/ha)","count:   2<br />value:  3.029330e+01<br />name: Pesticide intensity (EUR/ha)","count:   0<br />value:  3.304723e+01<br />name: Pesticide intensity (EUR/ha)","count:   0<br />value:  3.580117e+01<br />name: Pesticide intensity (EUR/ha)","count:   0<br />value:  3.855511e+01<br />name: Pesticide intensity (EUR/ha)","count:   2<br />value:  4.130904e+01<br />name: Pesticide intensity (EUR/ha)","count:   1<br />value:  4.406298e+01<br />name: Pesticide intensity (EUR/ha)","count:   0<br />value:  4.681691e+01<br />name: Pesticide intensity (EUR/ha)","count:   0<br />value:  4.957085e+01<br />name: Pesticide intensity (EUR/ha)","count:   1<br />value:  5.232479e+01<br />name: Pesticide intensity (EUR/ha)","count:   1<br />value:  5.507872e+01<br />name: Pesticide intensity (EUR/ha)","count:   1<br />value:  5.783266e+01<br />name: Pesticide intensity (EUR/ha)"],"type":"bar","textposition":"none","marker":{"autocolorscale":false,"color":"rgba(0,191,196,1)","line":{"width":1.88976377952756,"color":"transparent"}},"name":"Pesticide intensity (EUR/ha)","legendgroup":"Pesticide intensity (EUR/ha)","showlegend":true,"xaxis":"x4","yaxis":"y4","hoverinfo":"text","frame":null},{"orientation":"v","width":[0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669,0.175858568560669],"base":[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"x":[-1.58272711704602,-1.40686854848535,-1.23100997992468,-1.05515141136401,-0.879292842803343,-0.703434274242674,-0.527575705682006,-0.351717137121337,-0.175858568560669,0,0.175858568560669,0.351717137121337,0.527575705682006,0.703434274242674,0.879292842803343,1.05515141136401,1.23100997992468,1.40686854848535,1.58272711704602,1.75858568560669,1.93444425416735,2.11030282272802,2.28616139128869,2.46201995984936,2.63787852841003,2.8137370969707,2.98959566553137,3.16545423409204,3.3413128026527,3.51717137121337],"y":[2,4,12,26,47,61,76,60,80,78,71,107,128,140,185,215,234,262,237,191,185,135,93,51,25,19,5,3,0,2],"text":["count:   2<br />value: -1.582727e+00<br />name: Land-use diversity (0-100)","count:   4<br />value: -1.406869e+00<br />name: Land-use diversity (0-100)","count:  12<br />value: -1.231010e+00<br />name: Land-use diversity (0-100)","count:  26<br />value: -1.055151e+00<br />name: Land-use diversity (0-100)","count:  47<br />value: -8.792928e-01<br />name: Land-use diversity (0-100)","count:  61<br />value: -7.034343e-01<br />name: Land-use diversity (0-100)","count:  76<br />value: -5.275757e-01<br />name: Land-use diversity (0-100)","count:  60<br />value: -3.517171e-01<br />name: Land-use diversity (0-100)","count:  80<br />value: -1.758586e-01<br />name: Land-use diversity (0-100)","count:  78<br />value:  0.000000e+00<br />name: Land-use diversity (0-100)","count:  71<br />value:  1.758586e-01<br />name: Land-use diversity (0-100)","count: 107<br />value:  3.517171e-01<br />name: Land-use diversity (0-100)","count: 128<br />value:  5.275757e-01<br />name: Land-use diversity (0-100)","count: 140<br />value:  7.034343e-01<br />name: Land-use diversity (0-100)","count: 185<br />value:  8.792928e-01<br />name: Land-use diversity (0-100)","count: 215<br />value:  1.055151e+00<br />name: Land-use diversity (0-100)","count: 234<br />value:  1.231010e+00<br />name: Land-use diversity (0-100)","count: 262<br />value:  1.406869e+00<br />name: Land-use diversity (0-100)","count: 237<br />value:  1.582727e+00<br />name: Land-use diversity (0-100)","count: 191<br />value:  1.758586e+00<br />name: Land-use diversity (0-100)","count: 185<br />value:  1.934444e+00<br />name: Land-use diversity (0-100)","count: 135<br />value:  2.110303e+00<br />name: Land-use diversity (0-100)","count:  93<br />value:  2.286161e+00<br />name: Land-use diversity (0-100)","count:  51<br />value:  2.462020e+00<br />name: Land-use diversity (0-100)","count:  25<br />value:  2.637879e+00<br />name: Land-use diversity (0-100)","count:  19<br />value:  2.813737e+00<br />name: Land-use diversity (0-100)","count:   5<br />value:  2.989596e+00<br />name: Land-use diversity (0-100)","count:   3<br />value:  3.165454e+00<br />name: Land-use diversity (0-100)","count:   0<br />value:  3.341313e+00<br />name: Land-use diversity (0-100)","count:   2<br />value:  3.517171e+00<br />name: Land-use diversity (0-100)"],"type":"bar","textposition":"none","marker":{"autocolorscale":false,"color":"rgba(199,124,255,1)","line":{"width":1.88976377952756,"color":"transparent"}},"name":"Land-use diversity (0-100)","legendgroup":"Land-use diversity (0-100)","showlegend":true,"xaxis":"x3","yaxis":"y3","hoverinfo":"text","frame":null},{"x":[-9.36973806565879,-9.36973806565879],"y":[-16.9,354.9],"text":"mean: -9.369738","type":"scatter","mode":"lines","line":{"width":1.88976377952756,"color":"rgba(255,0,0,1)","dash":"dot"},"hoveron":"points","showlegend":false,"xaxis":"x","yaxis":"y","hoverinfo":"text","frame":null},{"x":[3.57396032213019,3.57396032213019],"y":[-32.15,675.15],"text":"mean:  3.573960","type":"scatter","mode":"lines","line":{"width":1.88976377952756,"color":"rgba(255,0,0,1)","dash":"dot"},"hoveron":"points","showlegend":false,"xaxis":"x2","yaxis":"y2","hoverinfo":"text","frame":null},{"x":[1.05852106117891,1.05852106117891],"y":[-13.1,275.1],"text":"mean:  1.058521","type":"scatter","mode":"lines","line":{"width":1.88976377952756,"color":"rgba(255,0,0,1)","dash":"dot"},"hoveron":"points","showlegend":false,"xaxis":"x3","yaxis":"y3","hoverinfo":"text","frame":null},{"x":[-1.40892435142209,-1.40892435142209],"y":[-42.3,888.3],"text":"mean: -1.408924","type":"scatter","mode":"lines","line":{"width":1.88976377952756,"color":"rgba(255,0,0,1)","dash":"dot"},"hoveron":"points","showlegend":false,"xaxis":"x4","yaxis":"y4","hoverinfo":"text","frame":null}],"layout":{"margin":{"t":47.0834371108344,"r":11.9551681195517,"b":60.9713574097136,"l":70.5354919053549},"plot_bgcolor":"rgba(255,255,255,1)","paper_bgcolor":"rgba(255,255,255,1)","font":{"color":"rgba(0,0,0,1)","family":"","size":23.9103362391034},"xaxis":{"domain":[0,0.441399945011402],"automargin":true,"type":"linear","autorange":false,"range":[-34.0632113593121,22.1410873835528],"tickmode":"array","ticktext":["-30","-20","-10","0","10","20"],"tickvals":[-30,-20,-10,0,10,20],"categoryorder":"array","categoryarray":["-30","-20","-10","0","10","20"],"nticks":null,"ticks":"outside","tickcolor":"rgba(51,51,51,1)","ticklen":5.97758405977584,"tickwidth":1.08683346541379,"showticklabels":true,"tickfont":{"color":"rgba(77,77,77,1)","family":"","size":19.1282689912827},"tickangle":-0,"showline":false,"linecolor":null,"linewidth":0,"showgrid":true,"gridcolor":"rgba(235,235,235,1)","gridwidth":1.08683346541379,"zeroline":false,"anchor":"y","title":"","hoverformat":".2f"},"annotations":[{"text":"value","x":0.5,"y":0,"showarrow":false,"ax":0,"ay":0,"font":{"color":"rgba(0,0,0,1)","family":"","size":23.9103362391034},"xref":"paper","yref":"paper","textangle":-0,"xanchor":"center","yanchor":"top","annotationType":"axis","yshift":-35.865504358655},{"text":"count","x":0,"y":0.5,"showarrow":false,"ax":0,"ay":0,"font":{"color":"rgba(0,0,0,1)","family":"","size":23.9103362391034},"xref":"paper","yref":"paper","textangle":-90,"xanchor":"right","yanchor":"center","annotationType":"axis","xshift":-45.4296388542964},{"text":"Fertilizer intensity (EUR/ha)","x":0.220699972505701,"y":1,"showarrow":false,"ax":0,"ay":0,"font":{"color":"rgba(26,26,26,1)","family":"","size":19.1282689912827},"xref":"paper","yref":"paper","textangle":-0,"xanchor":"center","yanchor":"bottom"},{"text":"GHG emissions (t)","x":0.779300027494299,"y":1,"showarrow":false,"ax":0,"ay":0,"font":{"color":"rgba(26,26,26,1)","family":"","size":19.1282689912827},"xref":"paper","yref":"paper","textangle":-0,"xanchor":"center","yanchor":"bottom"},{"text":"Land-use diversity (0-100)","x":0.220699972505701,"y":0.384310459154793,"showarrow":false,"ax":0,"ay":0,"font":{"color":"rgba(26,26,26,1)","family":"","size":19.1282689912827},"xref":"paper","yref":"paper","textangle":-0,"xanchor":"center","yanchor":"bottom"},{"text":"Pesticide intensity (EUR/ha)","x":0.779300027494299,"y":0.384310459154793,"showarrow":false,"ax":0,"ay":0,"font":{"color":"rgba(26,26,26,1)","family":"","size":19.1282689912827},"xref":"paper","yref":"paper","textangle":-0,"xanchor":"center","yanchor":"bottom"}],"yaxis":{"domain":[0.615689540845207,1],"automargin":true,"type":"linear","autorange":false,"range":[-16.9,354.9],"tickmode":"array","ticktext":["0","100","200","300"],"tickvals":[0,100,200,300],"categoryorder":"array","categoryarray":["0","100","200","300"],"nticks":null,"ticks":"outside","tickcolor":"rgba(51,51,51,1)","ticklen":5.97758405977584,"tickwidth":1.08683346541379,"showticklabels":true,"tickfont":{"color":"rgba(77,77,77,1)","family":"","size":19.1282689912827},"tickangle":-0,"showline":false,"linecolor":null,"linewidth":0,"showgrid":true,"gridcolor":"rgba(235,235,235,1)","gridwidth":1.08683346541379,"zeroline":false,"anchor":"x","title":"","hoverformat":".2f"},"shapes":[{"type":"rect","fillcolor":"transparent","line":{"color":"rgba(51,51,51,1)","width":1.08683346541379,"linetype":"solid"},"yref":"paper","xref":"paper","x0":0,"x1":0.441399945011402,"y0":0.615689540845207,"y1":1},{"type":"rect","fillcolor":"rgba(217,217,217,1)","line":{"color":"rgba(51,51,51,1)","width":1.08683346541379,"linetype":"solid"},"yref":"paper","xref":"paper","x0":0,"x1":0.441399945011402,"y0":0,"y1":38.2565379825654,"yanchor":1,"ysizemode":"pixel"},{"type":"rect","fillcolor":"transparent","line":{"color":"rgba(51,51,51,1)","width":1.08683346541379,"linetype":"solid"},"yref":"paper","xref":"paper","x0":0.558600054988598,"x1":1,"y0":0.615689540845207,"y1":1},{"type":"rect","fillcolor":"rgba(217,217,217,1)","line":{"color":"rgba(51,51,51,1)","width":1.08683346541379,"linetype":"solid"},"yref":"paper","xref":"paper","x0":0.558600054988598,"x1":1,"y0":0,"y1":38.2565379825654,"yanchor":1,"ysizemode":"pixel"},{"type":"rect","fillcolor":"transparent","line":{"color":"rgba(51,51,51,1)","width":1.08683346541379,"linetype":"solid"},"yref":"paper","xref":"paper","x0":0,"x1":0.441399945011402,"y0":0,"y1":0.384310459154793},{"type":"rect","fillcolor":"rgba(217,217,217,1)","line":{"color":"rgba(51,51,51,1)","width":1.08683346541379,"linetype":"solid"},"yref":"paper","xref":"paper","x0":0,"x1":0.441399945011402,"y0":0,"y1":38.2565379825654,"yanchor":0.384310459154793,"ysizemode":"pixel"},{"type":"rect","fillcolor":"transparent","line":{"color":"rgba(51,51,51,1)","width":1.08683346541379,"linetype":"solid"},"yref":"paper","xref":"paper","x0":0.558600054988598,"x1":1,"y0":0,"y1":0.384310459154793},{"type":"rect","fillcolor":"rgba(217,217,217,1)","line":{"color":"rgba(51,51,51,1)","width":1.08683346541379,"linetype":"solid"},"yref":"paper","xref":"paper","x0":0.558600054988598,"x1":1,"y0":0,"y1":38.2565379825654,"yanchor":0.384310459154793,"ysizemode":"pixel"}],"xaxis2":{"type":"linear","autorange":false,"range":[-44.4297780707897,68.3535047242918],"tickmode":"array","ticktext":["-25","0","25","50"],"tickvals":[-25,0,25,50],"categoryorder":"array","categoryarray":["-25","0","25","50"],"nticks":null,"ticks":"outside","tickcolor":"rgba(51,51,51,1)","ticklen":5.97758405977584,"tickwidth":1.08683346541379,"showticklabels":true,"tickfont":{"color":"rgba(77,77,77,1)","family":"","size":19.1282689912827},"tickangle":-0,"showline":false,"linecolor":null,"linewidth":0,"showgrid":true,"domain":[0.558600054988598,1],"gridcolor":"rgba(235,235,235,1)","gridwidth":1.08683346541379,"zeroline":false,"anchor":"y2","title":"","hoverformat":".2f"},"yaxis2":{"type":"linear","autorange":false,"range":[-32.15,675.15],"tickmode":"array","ticktext":["0","200","400","600"],"tickvals":[0,200,400,600],"categoryorder":"array","categoryarray":["0","200","400","600"],"nticks":null,"ticks":"outside","tickcolor":"rgba(51,51,51,1)","ticklen":5.97758405977584,"tickwidth":1.08683346541379,"showticklabels":true,"tickfont":{"color":"rgba(77,77,77,1)","family":"","size":19.1282689912827},"tickangle":-0,"showline":false,"linecolor":null,"linewidth":0,"showgrid":true,"domain":[0.615689540845207,1],"gridcolor":"rgba(235,235,235,1)","gridwidth":1.08683346541379,"zeroline":false,"anchor":"x2","title":"","hoverformat":".2f"},"xaxis3":{"type":"linear","autorange":false,"range":[-1.93444425416735,3.86888850833471],"tickmode":"array","ticktext":["-1","0","1","2","3"],"tickvals":[-1,0,1,2,3],"categoryorder":"array","categoryarray":["-1","0","1","2","3"],"nticks":null,"ticks":"outside","tickcolor":"rgba(51,51,51,1)","ticklen":5.97758405977584,"tickwidth":1.08683346541379,"showticklabels":true,"tickfont":{"color":"rgba(77,77,77,1)","family":"","size":19.1282689912827},"tickangle":-0,"showline":false,"linecolor":null,"linewidth":0,"showgrid":true,"domain":[0,0.441399945011402],"gridcolor":"rgba(235,235,235,1)","gridwidth":1.08683346541379,"zeroline":false,"anchor":"y3","title":"","hoverformat":".2f"},"yaxis3":{"type":"linear","autorange":false,"range":[-13.1,275.1],"tickmode":"array","ticktext":["0","100","200"],"tickvals":[-1.77635683940025e-15,100,200],"categoryorder":"array","categoryarray":["0","100","200"],"nticks":null,"ticks":"outside","tickcolor":"rgba(51,51,51,1)","ticklen":5.97758405977584,"tickwidth":1.08683346541379,"showticklabels":true,"tickfont":{"color":"rgba(77,77,77,1)","family":"","size":19.1282689912827},"tickangle":-0,"showline":false,"linecolor":null,"linewidth":0,"showgrid":true,"domain":[0,0.384310459154793],"gridcolor":"rgba(235,235,235,1)","gridwidth":1.08683346541379,"zeroline":false,"anchor":"x3","title":"","hoverformat":".2f"},"xaxis4":{"type":"linear","autorange":false,"range":[-27.5393613252448,63.340531048063],"tickmode":"array","ticktext":["-20","0","20","40","60"],"tickvals":[-20,0,20,40,60],"categoryorder":"array","categoryarray":["-20","0","20","40","60"],"nticks":null,"ticks":"outside","tickcolor":"rgba(51,51,51,1)","ticklen":5.97758405977584,"tickwidth":1.08683346541379,"showticklabels":true,"tickfont":{"color":"rgba(77,77,77,1)","family":"","size":19.1282689912827},"tickangle":-0,"showline":false,"linecolor":null,"linewidth":0,"showgrid":true,"domain":[0.558600054988598,1],"gridcolor":"rgba(235,235,235,1)","gridwidth":1.08683346541379,"zeroline":false,"anchor":"y4","title":"","hoverformat":".2f"},"yaxis4":{"type":"linear","autorange":false,"range":[-42.3,888.3],"tickmode":"array","ticktext":["0","200","400","600","800"],"tickvals":[0,200,400,600,800],"categoryorder":"array","categoryarray":["0","200","400","600","800"],"nticks":null,"ticks":"outside","tickcolor":"rgba(51,51,51,1)","ticklen":5.97758405977584,"tickwidth":1.08683346541379,"showticklabels":true,"tickfont":{"color":"rgba(77,77,77,1)","family":"","size":19.1282689912827},"tickangle":-0,"showline":false,"linecolor":null,"linewidth":0,"showgrid":true,"domain":[0,0.384310459154793],"gridcolor":"rgba(235,235,235,1)","gridwidth":1.08683346541379,"zeroline":false,"anchor":"x4","title":"","hoverformat":".2f"},"showlegend":false,"legend":{"bgcolor":"rgba(255,255,255,1)","bordercolor":"transparent","borderwidth":3.09234073013601,"font":{"color":"rgba(0,0,0,1)","family":"","size":19.1282689912827}},"hovermode":"closest","barmode":"relative"},"config":{"doubleClick":"reset","modeBarButtonsToAdd":["hoverclosest","hovercompare"],"showSendToCloud":false},"source":"A","attrs":{"43286dd6465e":{"x":{},"fill":{},"type":"bar"},"4328740713a4":{"xintercept":{}}},"cur_data":"43286dd6465e","visdat":{"43286dd6465e":["function (y) ","x"],"4328740713a4":["function (y) ","x"]},"highlight":{"on":"plotly_click","persistent":false,"dynamic":false,"selectize":false,"opacityDim":0.2,"selected":{"opacity":1},"debounce":0},"shinyEvents":["plotly_hover","plotly_click","plotly_selected","plotly_relayout","plotly_brushed","plotly_brushing","plotly_clickannotation","plotly_doubleclick","plotly_deselect","plotly_afterplot","plotly_sunburstclick"],"base_url":"https://plot.ly"},"evals":[],"jsHooks":[]}</script>

---
# Results: Heterogeneous impacts and statistical significance

<div id="nsiribfird" style="overflow-x:auto;overflow-y:auto;width:auto;height:auto;">
<style>@import url("https://fonts.googleapis.com/css2?family=Anton:ital,wght@0,100;0,200;0,300;0,400;0,500;0,600;0,700;0,800;0,900;1,100;1,200;1,300;1,400;1,500;1,600;1,700;1,800;1,900&display=swap");
html {
  font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, Oxygen, Ubuntu, Cantarell, 'Helvetica Neue', 'Fira Sans', 'Droid Sans', Arial, sans-serif;
}

#nsiribfird .gt_table {
  display: table;
  border-collapse: collapse;
  margin-left: auto;
  margin-right: auto;
  color: #333333;
  font-size: 16px;
  font-weight: normal;
  font-style: normal;
  background-color: #FFFFFF;
  width: auto;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #A8A8A8;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #A8A8A8;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
}

#nsiribfird .gt_heading {
  background-color: #FFFFFF;
  text-align: center;
  border-bottom-color: #FFFFFF;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
}

#nsiribfird .gt_title {
  color: #333333;
  font-size: 125%;
  font-weight: initial;
  padding-top: 4px;
  padding-bottom: 4px;
  padding-left: 5px;
  padding-right: 5px;
  border-bottom-color: #FFFFFF;
  border-bottom-width: 0;
}

#nsiribfird .gt_subtitle {
  color: #333333;
  font-size: 85%;
  font-weight: initial;
  padding-top: 0;
  padding-bottom: 6px;
  padding-left: 5px;
  padding-right: 5px;
  border-top-color: #FFFFFF;
  border-top-width: 0;
}

#nsiribfird .gt_bottom_border {
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#nsiribfird .gt_col_headings {
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
}

#nsiribfird .gt_col_heading {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: normal;
  text-transform: inherit;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: bottom;
  padding-top: 5px;
  padding-bottom: 6px;
  padding-left: 5px;
  padding-right: 5px;
  overflow-x: hidden;
}

#nsiribfird .gt_column_spanner_outer {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: normal;
  text-transform: inherit;
  padding-top: 0;
  padding-bottom: 0;
  padding-left: 4px;
  padding-right: 4px;
}

#nsiribfird .gt_column_spanner_outer:first-child {
  padding-left: 0;
}

#nsiribfird .gt_column_spanner_outer:last-child {
  padding-right: 0;
}

#nsiribfird .gt_column_spanner {
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  vertical-align: bottom;
  padding-top: 5px;
  padding-bottom: 5px;
  overflow-x: hidden;
  display: inline-block;
  width: 100%;
}

#nsiribfird .gt_group_heading {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: middle;
}

#nsiribfird .gt_empty_group_heading {
  padding: 0.5px;
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  vertical-align: middle;
}

#nsiribfird .gt_from_md > :first-child {
  margin-top: 0;
}

#nsiribfird .gt_from_md > :last-child {
  margin-bottom: 0;
}

#nsiribfird .gt_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  margin: 10px;
  border-top-style: solid;
  border-top-width: 1px;
  border-top-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 1px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 1px;
  border-right-color: #D3D3D3;
  vertical-align: middle;
  overflow-x: hidden;
}

#nsiribfird .gt_stub {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-right-style: solid;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  padding-left: 5px;
  padding-right: 5px;
}

#nsiribfird .gt_stub_row_group {
  color: #333333;
  background-color: #FFFFFF;
  font-size: 100%;
  font-weight: initial;
  text-transform: inherit;
  border-right-style: solid;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
  padding-left: 5px;
  padding-right: 5px;
  vertical-align: top;
}

#nsiribfird .gt_row_group_first td {
  border-top-width: 2px;
}

#nsiribfird .gt_summary_row {
  color: #333333;
  background-color: #FFFFFF;
  text-transform: inherit;
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
}

#nsiribfird .gt_first_summary_row {
  border-top-style: solid;
  border-top-color: #D3D3D3;
}

#nsiribfird .gt_first_summary_row.thick {
  border-top-width: 2px;
}

#nsiribfird .gt_last_summary_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#nsiribfird .gt_grand_summary_row {
  color: #333333;
  background-color: #FFFFFF;
  text-transform: inherit;
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
}

#nsiribfird .gt_first_grand_summary_row {
  padding-top: 8px;
  padding-bottom: 8px;
  padding-left: 5px;
  padding-right: 5px;
  border-top-style: double;
  border-top-width: 6px;
  border-top-color: #D3D3D3;
}

#nsiribfird .gt_striped {
  background-color: rgba(128, 128, 128, 0.05);
}

#nsiribfird .gt_table_body {
  border-top-style: solid;
  border-top-width: 2px;
  border-top-color: #D3D3D3;
  border-bottom-style: solid;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
}

#nsiribfird .gt_footnotes {
  color: #333333;
  background-color: #FFFFFF;
  border-bottom-style: none;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
}

#nsiribfird .gt_footnote {
  margin: 0px;
  font-size: 90%;
  padding-left: 4px;
  padding-right: 4px;
  padding-left: 5px;
  padding-right: 5px;
}

#nsiribfird .gt_sourcenotes {
  color: #333333;
  background-color: #FFFFFF;
  border-bottom-style: none;
  border-bottom-width: 2px;
  border-bottom-color: #D3D3D3;
  border-left-style: none;
  border-left-width: 2px;
  border-left-color: #D3D3D3;
  border-right-style: none;
  border-right-width: 2px;
  border-right-color: #D3D3D3;
}

#nsiribfird .gt_sourcenote {
  font-size: 90%;
  padding-top: 4px;
  padding-bottom: 4px;
  padding-left: 5px;
  padding-right: 5px;
}

#nsiribfird .gt_left {
  text-align: left;
}

#nsiribfird .gt_center {
  text-align: center;
}

#nsiribfird .gt_right {
  text-align: right;
  font-variant-numeric: tabular-nums;
}

#nsiribfird .gt_font_normal {
  font-weight: normal;
}

#nsiribfird .gt_font_bold {
  font-weight: bold;
}

#nsiribfird .gt_font_italic {
  font-style: italic;
}

#nsiribfird .gt_super {
  font-size: 65%;
}

#nsiribfird .gt_footnote_marks {
  font-style: italic;
  font-weight: normal;
  font-size: 75%;
  vertical-align: 0.4em;
}

#nsiribfird .gt_asterisk {
  font-size: 100%;
  vertical-align: 0;
}

#nsiribfird .gt_slash_mark {
  font-size: 0.7em;
  line-height: 0.7em;
  vertical-align: 0.15em;
}

#nsiribfird .gt_fraction_numerator {
  font-size: 0.6em;
  line-height: 0.6em;
  vertical-align: 0.45em;
}

#nsiribfird .gt_fraction_denominator {
  font-size: 0.6em;
  line-height: 0.6em;
  vertical-align: -0.05em;
}
</style>
<table class="gt_table" style="table-layout: fixed;; width: 0px">
  <colgroup>
    <col style="width:170px;"/>
    <col style="width:170px;"/>
    <col style="width:170px;"/>
    <col style="width:170px;"/>
    <col style="width:170px;"/>
  </colgroup>
  
  <thead class="gt_col_headings">
    <tr>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="color: #000000; font-family: Anton; font-size: small; text-align: left; vertical-align: middle;"></th>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 14px;">GHG emissions</th>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 14px;">Fertilizer intensity</th>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 14px;">Pesticide intensity</th>
      <th class="gt_col_heading gt_columns_bottom_border gt_left" rowspan="1" colspan="1" style="font-weight: bold;font-variant: small-caps;font-size: 14px;">Land use diversity</th>
    </tr>
  </thead>
  <tbody class="gt_table_body">
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px'>Share of (in-)significant AES participation effects
</div>
<div><span style ='font-weight:bold;color:darkgray;font-size:10px'>         (evaluated at the 95% significance level): 
         </span></div>
<div><span style ='font-weight:bold;color:navy;font-size:10px'>increase,</span>
<span style ='font-weight:bold;color:firebrick;font-size:10px'>decrease,</span>
<span style ='font-weight:bold;color:grey;font-size:10px'>insignificant</span></div></td>
<td class="gt_row gt_left"><img src="data:image/png;base64,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" style="height:168px;"></td>
<td class="gt_row gt_left"><img src="data:image/png;base64,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" style="height:168px;"></td>
<td class="gt_row gt_left"><img src="data:image/png;base64,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" style="height:168px;"></td>
<td class="gt_row gt_left"><img src="data:image/png;base64,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" style="height:168px;"></td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px'>Distributions of significant treatment effects
</div>
<div><span style ='font-weight:bold;color:gray;font-size:10px'>         (evaluated at the 95% significance level):</span></div></td>
<td class="gt_row gt_left"><img src="data:image/png;base64,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" style="height:168px;"></td>
<td class="gt_row gt_left"><img src="data:image/png;base64,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" style="height:168px;"></td>
<td class="gt_row gt_left"><img src="data:image/png;base64,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" style="height:168px;"></td>
<td class="gt_row gt_left"><img src="data:image/png;base64,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" style="height:168px;"></td></tr>
    <tr><td class="gt_row gt_left"><div><span style='font-weight:bold;font-variant:small-caps;font-size:14px'>Mean significant effects</div>
<div><span style ='font-weight:bold;color:grey;font-size:10px'>(evaluated at the 95% significance level)</span></div></td>
<td class="gt_row gt_left"><div><span style='color:navy;font-size:14px'>Mean(+): 12t</div>
<div><span style='color:firebrick;font-size:14px'>Mean(-): -10.8t</div></td>
<td class="gt_row gt_left"><div><span style='color:navy;font-size:14px'>Mean(+): /</div>
<div><span style='color:firebrick;font-size:14px'>Mean(-): -14.3EUR/ha</div></td>
<td class="gt_row gt_left"><div><span style='color:navy;font-size:14px'>Mean(+): 6.6EUR/ha</div>
<div><span style='color:firebrick;font-size:14px'>Mean(-): -10.3EUR/ha</div></td>
<td class="gt_row gt_left"><div><span style='color:navy;font-size:14px'>Mean(+): 1.6</div>
<div><span style='color:firebrick;font-size:14px'>Mean(-): -0.9</div></td></tr>
  </tbody>
  
  
</table>
</div>

---
# Results: Additionality, adversity and windfall effects
.left-column[
<br/>
- We find additionality in approx. 70% of the observations

- We find adverse effects in approx. 6% of the observations

- We find windfall effects in approx. 30% of the observations
]

.right-column[
.pull-left[
<img src="data:image/png;base64,#img/map_additional.png" width="100%" />
]

.pull-right[
<img src="data:image/png;base64,#img/map_adverse.png" width="100%" />
]]

---
# Things I have not talked about...

<br/>

**...how we tackled potential sources of endogeneity**

- Selection-on-observables
- Bad controls problem
- Unobserved confounding

<br/>

**...the robustness of our results**

- i.e. robustness and sensitivity analyses we conducted

<br/>

**... how we used model-agnostic interpretation for improved scheme targeting**

---
# Major limitations of the study
<br/>

- Indicator choice

<br/>

- Effect heterogeneity vs. treatment heterogeneity

<br/>

- Cross-sectional nature of the study & the possibility of unobserved confounding

---
name: conclusions
# Conclusions

<br/>

- Generalized random forest appears to be a good addition to the agricultural subsidies impact assessment toolbox

<br/>

- We find heterogeneous impacts of AES participation

<br/>

- AES are not very effective in terms of the chosen environmental indicators

<br/>

- Ignoring impact heterogeneity yields an incomplete picture of the AES effectiveness

<br/>

**Our analysis should be seen as a first step in the direction of a more complete
assessment of agri-environmental schemes.**

---
class: hide-count, middle center
count: false
# Thank you!
[References](#references)

[Appendices](#appendix)

---
class: hide-count
count: false
name: references
# References

<a name=bib-Uehleke2022></a>[[1]](#cite-Uehleke2022) R. Uehleke, M.
Petrick, and S. Hüttel. "Evaluations of agri-environmental schemes
based on observational farm data: The importance of covariate
selection". In: _Land Use Policy_ 114 (2022), p. 105950.

<a name=bib-Pufahl2009></a>[[2]](#cite-Pufahl2009) A. Pufahl and C. R.
Weiss. "Evaluating the effects of farm programmes: Results from
propensity score matching". In: _European Review of Agricultural
Economics_ 36.1 (2009), pp. 79-101.

<a name=bib-Arata2016></a>[[3]](#cite-Arata2016) L. Arata and P.
Sckokai. "The impact of agri-environmental schemes on farm performance
in five E.U. member States: A DID-matching approach". In: _Land
Economics_ 92.1 (2016), pp. 167-186.

<a name=bib-Bertoni2020></a>[[4]](#cite-Bertoni2020) D. Bertoni, D.
Curzi, G. Aletti, et al. "Estimating the effects of agri-environmental
measures using difference-in-difference coarsened exact matching". In:
_Food Policy_ 90 (2020), p. 101790.

<a name=bib-Wuepper2021></a>[[5]](#cite-Wuepper2021) D. Wuepper and R.
Huber. " Comparing effectiveness and return on investment of action-
and results-based agri-environmental payments in Switzerland ". In:
_American Journal of Agricultural Economics_ (2021), pp. 1-20.

<a name=bib-Athey2019f></a>[[6]](#cite-Athey2019f) S. Athey, J.
Tibshirani, and S. Wager. "Generalized random forests". In: _Annals of
Statistics_ 47.2 (2019), pp. 1179-1203. eprint: 1610.01271.

---
count: false
name: appendix

# Appendices

1. [Generalized random forests](#app_grf)

1. [Explaining ML models](#app_interpretation)

1. [Shapley values](#app_shapley_value_explain)

1. [Results: Shapley values for targeting](#app_results_shapley)

1. [Endogeneity](#app_endogeneity)

1. [DAGs: Identification](#app_sensitivity)

1. [DAGs: Robustness tests](#app_robustness_explain)

1. [DAGs: Robustness test results](#app_robustness_result)

1. [Sensitivity analysis: unobserved confounding](#app_robustness_explain)

1. [An approximation to the cost-effectiveness](#cost-effectiveness)
---
name: app_grf
count: false
class: split-four
# Generalized random forests

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree1.png" width="95%" style="display: block; margin: auto 0 auto auto;" />
]

.column[

]

.column[

]

.column[

]

---
class: split-four
count: false
# Generalized random forests
.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree1.png" width="95%" style="display: block; margin: auto 0 auto auto;" />

<p style="margin-top:1.9cm; white-space: pre;">  Observation space<p>
<img src="data:image/png;base64,#img/weight1.png" width="90%" style="display: block; margin: auto 0 auto auto;" />
]

.column[

]

.column[

]

.column[

]

---
class: split-four
count: false
# Generalized random forests

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree1.png" width="95%" style="display: block; margin: auto 0 auto auto;" />

<p style="margin-top:1.9cm; white-space: pre;">  Observation space<p>
<img src="data:image/png;base64,#img/weight1.png" width="90%" style="display: block; margin: auto 0 auto auto;" />
]

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree2.png" width="82%" style="display: block; margin: auto;" />

<img src="data:image/png;base64,#img/weight2.png" width="90%" style="display: block; margin: auto;" />
]

.column[

]

.column[

]

---
class: split-four
count: false
# Generalized random forests

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree1.png" width="95%" style="display: block; margin: auto 0 auto auto;" />

<p style="margin-top:1.9cm; white-space: pre;">  Observation space<p>
<img src="data:image/png;base64,#img/weight1.png" width="90%" style="display: block; margin: auto 0 auto auto;" />
]

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree2.png" width="82%" style="display: block; margin: auto;" />

<img src="data:image/png;base64,#img/weight2.png" width="90%" style="display: block; margin: auto;" />
]

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree3.png" width="83%" />

<img src="data:image/png;base64,#img/weight3.png" width="90%" />
]

.column[

]

---
class: split-four
count: false
# Generalized random forests

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree1.png" width="95%" style="display: block; margin: auto 0 auto auto;" />

<p style="margin-top:1.9cm; white-space: pre;">  Observation space<p>
<img src="data:image/png;base64,#img/weight1.png" width="90%" style="display: block; margin: auto 0 auto auto;" />
]

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree2.png" width="82%" style="display: block; margin: auto;" />

<img src="data:image/png;base64,#img/weight2.png" width="90%" style="display: block; margin: auto;" />
]

.column[
<br/>
<br/>
<br/>
<br/>
<img src="data:image/png;base64,#img/tree3.png" width="83%" />

<img src="data:image/png;base64,#img/weight3.png" width="90%" />
]

.column[
<br/>
<br/>
<br/>
<br/>
<br/>
<br/>
.pull-left[
`$$\boxed{\hat{\tau}(x) = \\ \min_{\tau} \sum_{i=1}^{n} \boxed{ \color{red}{\alpha_i(x)}} \times \\\color{navy}{(Y_i - (const.+\tau W_i ))^2}}$$`]

.font80[[Overview appendix](#appendix)]
]

---
name: app_interpretation
count: false
# Shapley values: But how can we interpret the results?
## Making sense of the black box

.pull-left[
### GRF in practice
<img src="data:image/png;base64,#img/example_tree_ghg.png" width="100%" />
]

.pull-right[
### Interpreting ML models

- Trade-off between complexity and interpretability

- We need a mechanism to trace back where the heterogeneity comes from

- Explainable machine learning

- Model-agnostic methods

.font80[[Overview appendix](#appendix)]
]

---
name: app_shapley_value_explain
count: false
# Shapley values
+ Concept stems from cooperative game theory

+  Shapley values explain the contributions of different variables to an individual prediction.

+ The sum of Shapley values yields the difference of actual and average prediction.

+ It is NOT the difference in prediction when we would remove the feature from the model.

<br/>
<br/>
+ A simplistic low-dimensional regression analogy in our context: `$$GHG = \alpha + \beta_{1} Treatment + \color{red}{\beta_{12}}(Treatment \times \color{red}{\text{farm type)}} + ... + error$$`

.font80[[Overview appendix](#appendix)]

---
count: false
# Shapley values: Coalitions
<img src="data:image/png;base64,#img/stylized_shapley.png" width="100%" />

.font80[[Overview appendix](#appendix)]

---
count: false
# Shapley values: Global
<img src="data:image/png;base64,#img/stylized_shapley_global.png" width="100%" />

.font80[[Overview appendix](#appendix)]

---
name: app_results_shapley
count: false
# Results: Shapley values
.left-column[
<br/>
**Examplary contextual feature: location**

- How do contextual variables drive AES impact?

- E.g, location drives AES impact

- Targeting regions with high Shapley values could improve effectiveness

.font80[[Overview appendix](#appendix)]
]

.right-column[
<img src="data:image/png;base64,#img/gg_facet_shap_RegID_erae.png" width="110%" />
]

---
name: app_endogeneity
count: false
# Tackling Endogeneity

**Through estimation method (selection-on-observables)**

- Propensity score adjustment
    - Doubly robust estimator
    - Many control variables

**Through contextual variable adjustment (bad controls problem)**

- Cross-sectional analysis
    - 2007-2013 average values for contextual variables: pre-treatment

**Through algorithmic complexity (unobserved confounding)**

- Unobserved heterogeneity usually not contained in observables
    - Complex interactions and nonlinear combination of observables might contain
    latent variables

**We ran multiple sensitivity analyses, robustness checks, simulations of potential unobserved confounders**

.font80[[Overview appendix](#appendix)]

---
name: app_dag_endogeneity
count: false
name: conclusions

# DAG: Selection-on-observables
.pull-left[
<img src="data:image/png;base64,#img/dag1.png" width="200%" />
]

.pull-right[
Directed acyclic graph (DAG) without unobserved confounders, i.e. the unconfoundedness
assumption is fulﬁlled. The eﬀect of the treatment variable D (i.e. participation in AES)
on an outcome Y (i.e. environmental indicator) is identiﬁed if our model controls for all
observed confounders X 1 through X K (i.e. contextual variables) and hence all backdoor
paths are closed. Connections among confounders are not shown for brevity.

.font80[[Overview appendix](#appendix)]]

---
count: false
name: conclusions

# DAG: Latent confounders
.pull-left[
<img src="data:image/png;base64,#img/DAG2_unobs.png" width="200%" />
]

.pull-right[
.font80[
Directed acyclic graph (DAG) in an exemplary situation where two unobserved
confounders (U 1 , U 2 ) are present. The eﬀect of the treatment variable D (i.e.
participation in AES) on an outcome Y (i.e. environmental indicator) is not correctly
identiﬁed if our model does not control for all observed confounders X 1 through X K and
unobserved confounders (U 1 , U 2 ). Since U 1 and U 2 are not observable, there is no way to
directly control for these confounders. Yet, under the assumption that observed and
unobserved confounders are associated (arrows from U to X) and the unobserved
confounders are reﬂected in the complex, nonlinear, and high-dimensional combination of
the large number of observed confounders (latent confounder space), it might be possible
to capture (most of) the variation coming from the unobserved confounder space, if the
causal forest maps that latent confounder space accurately. If this is the case, the
backdoor path D ← U 1,2 → Y can be closed and the treatment eﬀect from AES
participation on environmental performance is identiﬁed. Connections among confounders
are not shown for brevity.]

.font80[[Overview appendix](#appendix)]]

---
count: false
name: app_robustness_explain

# Robustness: Random placebo treatment
.pull-left[
<img src="data:image/png;base64,#img/DAGr_Wrandom.png" width="200%" />
]

.pull-right[
.font80[
The original treatment variable (D) is replaced by a random placebo treatment variable ($D_{random}$). If the model is correctly speciﬁed, we expect no effect of the
treatment on the environmental outcome.]

.font80[[Overview appendix](#appendix)]]

---
count: false
name: conclusions
# Robustness: Random outcome
.pull-left[
<img src="data:image/png;base64,#img/DAGr_Yrandom.png" width="200%" />
]

.pull-right[
.font80[
 The original outcome variable (Y) is replaced by a random placebo outcome variable ($Y_{random}$). If the model is correctly speciﬁed, we expect no effect of the treatment on the environmental outcome.]

.font80[[Overview appendix](#appendix)]]

---
count: false
name: conclusions
# Robustness: Extra random covariate
.pull-left[
<img src="data:image/png;base64,#img/DAGr_Xrandom.png" width="200%" />
]

.pull-right[
.font80[
We add an extra random variable ($U_{random}$) as potential confounder to our model. As this confounder is random, we expect no eﬀect of the treatment on the environmental outcome.]

.font80[[Overview appendix](#appendix)]]

---
count: false
name: conclusions
# Robustness: leave out most important confounder(s)
.pull-left[
<img src="data:image/png;base64,#img/DAGr_LO_X.png" width="200%" />
]

.pull-right[
.font80[We leave out the (three) most important confounder(s) and re-estimate the model. If the (nonlinear) correlation structure of the other observed confounders properly reﬂect the left-out variable(s), we expect no change of the model against our baseline model, which includes all observed confounders. This could point towards the stability of our model against unobserved confounders that are not included in the model if they are associated
with the set of observed confounders.]

.font80[[Overview appendix](#appendix)]]

---
count: false
name: conclusions
# Robustness: Leave out confounder(s) based on PCA
.pull-left[
<img src="data:image/png;base64,#img/DAGr_LO_PC.png" width="200%" />
]

.pull-right[
.font80[We leave out various observed confounders more systematically. We use principal component analysis and the resulting loadings to detect systematic groups of confounders and
leave these out, such that the combination structure of the other observed confounders are less capable to compensate for the exclusion of these variables. The retrieved model behavior could indicate how strongly it reacts to potentially completely left out confounders that
are not buffered away by the combination of observed covariates.]

.font80[[Overview appendix](#appendix)]]

---
count: false
name: app_robustness_result
# Robustness: Results

.pull-left[
**GHG emissions**
<img src="data:image/png;base64,#img/compplot_ghg_robust.png" width="100%" />
]

.pull-right[
**Fertilizer intensity**
<img src="data:image/png;base64,#img/compplot_fert_robust.png" width="100%" />
]

.font80[[Overview appendix](#appendix)]

---
count: false
name: conclusions
# Robustness: Results

.pull-left[
**Pesticide intensity**
<img src="data:image/png;base64,#img/compplot_pest_robust.png" width="100%" />
]

.pull-right[
**Land-use diversity**
<img src="data:image/png;base64,#img/compplot_div_robust.png" width="100%" />
]

.font80[[Overview appendix](#appendix)]

---
count: false
name: app_sensitivity
# Sensitivity: Simulate unobserved confounder
.pull-left[
<img src="data:image/png;base64,#img/DAGr_CCtarget.png" width="200%" />
]

.pull-right[
.font80[In this robustness check, we simulate a completely left out unobserved covariate ($U_{target}$) simulating various diffrent correlation structures between Utarget, D and
Y. This gives indication as to how strong the omitted variable bias in our model could be given dfferent correlation structures between Utarget, D and Y.]

.font80[[Overview appendix](#appendix)]]

---
count: false
name: conclusions
# Sensitivity: Simulate unobserved confounder

.pull-left[
**GHG emissions**

<img src="data:image/png;base64,#img/gg_robust_ovb_ghg.png" width="65%" />
]

.pull-right[
**Fertilizer intensity**
<img src="data:image/png;base64,#img/gg_robust_ovb_fert.png" width="65%" />
]

.font80[[Overview appendix](#appendix)]
---
count: false
name: conclusions
# Sensitivity: Simulate unobserved confounder

.pull-left[
**Pesticide intensity**
<img src="data:image/png;base64,#img/gg_robust_ovb_pest.png" width="65%" />
]

.pull-right[
**Land-use diversity**
<img src="data:image/png;base64,#img/gg_robust_ovb_div.png" width="65%" />
]

.font80[[Overview appendix](#appendix)]

---
count: false
name: cost-effectiveness
# Results: An approximation to the cost-effectiveness of AES participation

.font80[[Overview appendix](#appendix)]