Periodic Reporting for period 1 - CLIMOD (Integrating climate impact and spatial microsimulation modelling for improved climate change adaptation decision-making)
Periodo di rendicontazione: 2018-04-02 al 2020-04-01
Minimising the welfare cost of climate change requires effective and efficient adaptation policy design. Accurate measurement of welfare cost is a prerequisite for effective policy prioritisation. Microdata-based analyses are important to understand the distribution of impacts between rich and poor and to incorporate these effects into policy decision-making. Of particular importance is the dearth of spatially-referenced microdata in many countries, limiting the insight one can glean in relation to the socioeconomic incidence of climate change impacts. As a result, much of the current state of the art quantifies welfare costs using aggregated methodologies. This has been shown to bias cost estimates and does not capture the socioeconomic distribution of welfare impacts. This bias may affect the effectiveness of adaptation measures. This project integrates climate impact modelling with Spatial Microsimulation (SM) and other micro-level socioeconomic modelling to estimate micro-level welfare impacts to improve the accuracy of adaptation decision-making.
This project had two primary research objectives; (1) the creation and testing of a new SM methodology that may widen the scope of application to a wider number of contexts and geographical regions; (2) the integration of this new methodology with climate impact modelling at PIK and (3) show how distributional impacts of climate change may be used to inform adaptation policy decision-making.
This project had two primary research objectives; (1) the creation and testing of a new SM methodology that may widen the scope of application to a wider number of contexts and geographical regions; (2) the integration of this new methodology with climate impact modelling at PIK and (3) show how distributional impacts of climate change may be used to inform adaptation policy decision-making.
The work performed may be divided into two sections, development of spatial microsimulation (SM) framework for improved climate impact analysis and the micro-level distributional evaluation of climate impacts. This resulted in three working papers. It should be noted that, with minor amendments, the three work packages of this action have been delivered despite the action ending 4 months early. Some minor amendments to WP3 were applied to allow for this time change.
The first goal of this project was to create a new method known as ‘Conditional Monte Carlo’ (CMC). This is a spatial microsimulation methodology to quantify spatially-explicit micro-level welfare analysis of climate impacts. This method is that it overcomes many data limitations presented by existing methods; it may be used for contexts where previous methods provided unreliable results. Further, the data burden is less onerous than other methods. This flexibility facilitated a robust validation procedure where Mexican socioeconomic data were used. Mexico is the only country for which data is collected at the small area level. This provides a most robust and credible benchmark of model performance, against which the procedure performed exceptionally well, simulating welfare with a degree of precision unprecedented in the literature. The attached figure provides insight into how well the procedure has reproduced the socioeconomic data. A working paper was produced. It is envisaged that this paper will be submitted to a 3* or 4-ranked economics journal on the ABS scale.
The second objective demonstrated the use of the SM model for climate impact analysis and adaptation decision-making. PIK climate scientists aided the PI, Dr. Farrell, with climate impact modelling and Dr. Farrell then proceeded to link this climate impact analysis with the socioeconomic data for the CMC procedure.
A working paper links socioeconomic data with climate impact data on floods and temperature extremes. It shows that current methods have aggregation bias and also shows the socioeconomic distributional impacts of climate change in Mexico, something previously untested in the literature. It is envisaged that this paper will be submitted to Nature Climate Change or a similar high-impact journal.
The third working paper assessed the impacts new technology adoption may have for food security and climate change adaptation. This was carried out in conjunction with climate scientists at PIK who aided in the linking of climate data with socioeconomic microdata. This paper shows that improved maize seed increases food availability in developing countries, potentially mitigating the negative effects of climate change, but the distribution of this benefit may not necessarily remedy food insecurity. Adoption is often biased towards well-resourced households. The corrections required for overcoming this bias can be identified through a novel analysis of the distribution of adoption determinants across the income spectrum, where targeting improved seed adoption towards households headed by older persons, uneducated, those with off-farm income and households headed by females yields equal adoption across the welfare spectrum.
This paper highlights the importance of looking beyond aggregate effects and to understand the distribution of impacts to fully understand the effectiveness of a climate adaptation intervention. This paper is currently in working paper format and will be submitted to The ‘American Journal of Agricultural Economics’ or a similarly high-impact journal.
All working papers have yet to pass peer review. Dissemination will occur after peer review, in line with common practice. Further public dissemination will be facilitated through podcast publication, a research bulletin and website dissemination, upon passing peer review. A podcast has been established through this action and this will feature dissemination of this work upon peer-review. This work has been submitted for presentation at academic conferences in 2020. International regional science and climate economics conferences have been targetted.
The first goal of this project was to create a new method known as ‘Conditional Monte Carlo’ (CMC). This is a spatial microsimulation methodology to quantify spatially-explicit micro-level welfare analysis of climate impacts. This method is that it overcomes many data limitations presented by existing methods; it may be used for contexts where previous methods provided unreliable results. Further, the data burden is less onerous than other methods. This flexibility facilitated a robust validation procedure where Mexican socioeconomic data were used. Mexico is the only country for which data is collected at the small area level. This provides a most robust and credible benchmark of model performance, against which the procedure performed exceptionally well, simulating welfare with a degree of precision unprecedented in the literature. The attached figure provides insight into how well the procedure has reproduced the socioeconomic data. A working paper was produced. It is envisaged that this paper will be submitted to a 3* or 4-ranked economics journal on the ABS scale.
The second objective demonstrated the use of the SM model for climate impact analysis and adaptation decision-making. PIK climate scientists aided the PI, Dr. Farrell, with climate impact modelling and Dr. Farrell then proceeded to link this climate impact analysis with the socioeconomic data for the CMC procedure.
A working paper links socioeconomic data with climate impact data on floods and temperature extremes. It shows that current methods have aggregation bias and also shows the socioeconomic distributional impacts of climate change in Mexico, something previously untested in the literature. It is envisaged that this paper will be submitted to Nature Climate Change or a similar high-impact journal.
The third working paper assessed the impacts new technology adoption may have for food security and climate change adaptation. This was carried out in conjunction with climate scientists at PIK who aided in the linking of climate data with socioeconomic microdata. This paper shows that improved maize seed increases food availability in developing countries, potentially mitigating the negative effects of climate change, but the distribution of this benefit may not necessarily remedy food insecurity. Adoption is often biased towards well-resourced households. The corrections required for overcoming this bias can be identified through a novel analysis of the distribution of adoption determinants across the income spectrum, where targeting improved seed adoption towards households headed by older persons, uneducated, those with off-farm income and households headed by females yields equal adoption across the welfare spectrum.
This paper highlights the importance of looking beyond aggregate effects and to understand the distribution of impacts to fully understand the effectiveness of a climate adaptation intervention. This paper is currently in working paper format and will be submitted to The ‘American Journal of Agricultural Economics’ or a similarly high-impact journal.
All working papers have yet to pass peer review. Dissemination will occur after peer review, in line with common practice. Further public dissemination will be facilitated through podcast publication, a research bulletin and website dissemination, upon passing peer review. A podcast has been established through this action and this will feature dissemination of this work upon peer-review. This work has been submitted for presentation at academic conferences in 2020. International regional science and climate economics conferences have been targetted.
First, the spatial microsimulation method progresses the state of the art as it suits a wider range of data with greater ease. It may be used for contexts where previous methods provided unreliable results. Further, the data burden is less onerous than other methods. This provides a tool that may aid climate adaptation. This will allow for spatially-disaggregated data to be created and aligned with spatially disaggregated climate data to better inform climate change adaptation policy.
Second, the benefits of spatial disaggregation in climate impact analysis has been demonstrated by this work. The biased estimates of welfare loss and inefficient adaptation policy that may arise from commonly employed aggregated methods have both been highlighted; using disaggregated data improves climate impact estimation and this project has demonstrated how this may be accurately simulated in the absence of such data.
The third contribution of this paper highlighted the importance of considering the distribution of impacts, and not just the aggregated impact, when considering climate adaptation and food security policy. The corrections required for overcoming this bias can be identified through a novel analysis of the distribution of adoption determinants across the income spectrum. Thus, recommended policy actions for greater food security in light of climate change have been identified. These are simple and implementable, the method of distribution should take into account underlying measures of welfare (income, assets, expenditure, etc.) in order to provide resilience to climate change, and socioeconomic proxies in the absence of this information have been identified.
To summarise, these contributions inform climate adaptation policy in fundamental ways, providing tools for analysis and additional insight using distributional methods.
Second, the benefits of spatial disaggregation in climate impact analysis has been demonstrated by this work. The biased estimates of welfare loss and inefficient adaptation policy that may arise from commonly employed aggregated methods have both been highlighted; using disaggregated data improves climate impact estimation and this project has demonstrated how this may be accurately simulated in the absence of such data.
The third contribution of this paper highlighted the importance of considering the distribution of impacts, and not just the aggregated impact, when considering climate adaptation and food security policy. The corrections required for overcoming this bias can be identified through a novel analysis of the distribution of adoption determinants across the income spectrum. Thus, recommended policy actions for greater food security in light of climate change have been identified. These are simple and implementable, the method of distribution should take into account underlying measures of welfare (income, assets, expenditure, etc.) in order to provide resilience to climate change, and socioeconomic proxies in the absence of this information have been identified.
To summarise, these contributions inform climate adaptation policy in fundamental ways, providing tools for analysis and additional insight using distributional methods.