Back in the summer last year, we found ourselves knee deep in waste. Thankfully a metaphor and not a real sanitation problem at the office. We were working on a fascinating yet complicated project for the Welsh Government—reviewing the impact of the Wales Waste Plan (WWP).
Towards Zero Waste (TZW) and the improvement it has heralded must surely must be one of Welsh Government’s greatest achievements of its 20 years in power—I attended a celebration of 20 years of Welsh Government last year and sadly the sobering conclusion of the party was that it has done little with its devolved power—and so the fact that Wales is currently ranked 2nd best in the world for household waste recycling should be lauded.
A key mechanism of implementing the strategy has been the Wales Waste Plan (WWP) and specifically the Sustainable Waste Management Grant (SWMG). That is, money given to Local Authorities and other contracted organisations to manage household, industrial & commercial (I and C), and construction, demolition and excavation (CD&E) waste—focused on a collective aim to reach a recycling target of 70% by 2025. In doing so, increasing recycling rates from an unbelievably low 5% when the Welsh Government first took office.
My role in the highly complex review was to design a way of measuring the economic impact. Basically, answering the question; How much money has the waste plan saved the Welsh economy? The answer demanded the creation of a bespoke economic impact model. I had to revisit that model a few weeks ago and in doing so I was struck by its complexity that I thought it deserved unpicking to find out what I’d learnt through the process. So, here’s my top four lessons for designing an economic impact model of the waste management industry.
- It’s complicated so break it down – the interconnected nature of any industry is complicated, and the waste industry is no exception. The source of potential economic benefit from waste reduction is particularly complex. There’s the impact of direct spend, the cost savings from improved efficiency, carbon cost reductions and mitigation costs avoided, supply chain effects and so on. Add into that the different funding streams, for different purposes and the picture is chaotic. The only way to build a model that could make sense of and capture all effects was to break it down into small chunks.
For the WWP there were several streams:
- The SWMG—given to local authorities to help them to deliver services and provide the necessary infrastructure to deliver national policy outcomes. The breadth of activities could be grouped into landfill avoidance, re-use and recycling.
- Carbon Savings – the economic benefits accruing from the savings made as a result of reducing carbon emissions through the activity of the WWP. Acknowledging that these are savings that would have been spent elsewhere and so displaced spend rather than net benefit.
- The impact of WRAP activity – WRAP Cymru receives a significant proportion of its funding from the Welsh Government, and its activities stretch far beyond that of the SWMG alone. The impact of their activities was captured in other studies, reviewed to inform this model.
Overview of the SWMG Economic Impact Assessment Approach
- Accept the limitations – models aren’t perfect. They can only infer what might be the case, it is terribly difficult to differentiate between impacts which overlap, so don’t expect them to do it all. Appreciate what can and what cannot be measured and stick to it. Explain why something can’t be measured. Then explain what can be measured and fully explain the limitations. For the WWP this was particularly challenging. The available data was fairly comprehensive, yet gaps existed. Trend analysis and assumptions had to be made to fill gaps in data and this is the main area where models differ. A slightly different assumption has significant impacts on output estimations. Take care over assumptions and be clear what they are and how they impact the model estimations. For the SWMG, initial spend data was available for part of the 15 year period under review. We modelled scenarios to estimate total spend. We then estimated leakage – how much was spent outside Wales, using published data on processed waste; before calculating impacts as discussed below.
- Don’t reinvent the wheel – there are so many high-quality resources that can provide the information you need. The waste sector is particularly full of excellent research stating the savings of various types of waste. For the SWMG, we modelled different strands: landfill avoidance, recyclate generated, and re-use. For each, various existing figures were used to estimate gross impact. This included deriving a price per tonne for each type of recyclate (there are many). GVA to turnover ratios (using National Datasets) were used to understand impact in each industry where spend is thought to (another assumption) occur and input/output tables referenced to understand the knock-on effect of spend (indirect and induced effects in economic parlance). A similar approach was used for the re-use of materials. For carbon savings, both savings from the WWP and the SWMG were modelled using a three-step approach. The first step was to define the scope of actions for which a carbon saving value could be obtained. Secondly, existing research was utilised to monetise the environmental effects of carbon savings so a savings value could be obtained. This step was vital in our model. The Government Economic Service Working Papers and other academic papers provided estimates of the cost of marginal damage per tonne of carbon, including year-on-year costs avoided (from less carbon being emitted into the environment). Estimates of costs for each waste category (landfill avoidance, I&C and CD&E) were then obtained using a few conversion ratios (the ratio of the atomic weight of carbon to carbon dioxide was particularly useful). The final step was to use a standard economic modelling approach to determine the contribution of carbon saving to GVA—considering additionality and accounting for general price rises (GDP deflators were used).
- Never underestimate the importance of a well-designed spreadsheet. This seems obvious but when you are in the throes of a project, fully immersed and everything makes sense in your head it takes great strength to take a step back and spend time designing your spreadsheet to be fully traceable, fully linked and fool-proof. It will save you days when the inevitable happens and a data input changes, or the price of a material changes. At the time I compiled the model I knew I was spending a lot of time on simply structuring the spreadsheet so that all cells were automated – referring to a clearly marked data input cell from which other cells were calculated with formulas. Again – I know it’s a fundamental aspect of effectively using spreadsheets, but it’s so tempting to put the number you know is right into a cell, when you should put it into a reference box instead. Thankfully I did this on the Waste Plan model, so when the Welsh Government updated their C&I spend nine months after it was first drafted, I just needed to change two cells and the model recalculated the answer. I could have cried with happiness.
I think these lessons carry across to any economic impact model design. I hope they help others who venture into the less than scientific world of economic impact modelling. I know I will keep these points front of mind the next time I’m tasked with the challenge. I just hope I can stick to them!
 Clarkson, R, and Deyes, K, ‘Estimating the Social Cost of Carbon Emissions’ January 2002. Available at [ftp://220.127.116.11/Transfer/ES_Pubs/ESVal/carbon_val/clarkson_02_socialCostCarbon_ukgov140.pdf].