Salinity conference

Risk and investment analysis
Dr Stefan Hajkowicz

Stefan Hajkowicz undertook his studies at the Department of Geographical Sciences and Planning at the University of Queensland, completing an undergraduate degree in 1995 and a doctoral thesis in 2000. He has since completed a postgraduate qualification in economics at the University of New England. In his current position with CSIRO, Stefan works on issues relating to environmental and natural resource management policy. He has a particular interests in decision support, spatial modelling and investment appraisal.

I am an environmental and resource economist. A lot of my reactions to the report are quite similar to what Andy has just brought up, and I think you will see some of those in my presentation here.

Click on image for a larger version of figure 1

What, obviously, was most relevant to me in the report was section 2.2, on investment decisions. As it see it, there are two bits of investment appraisal going on here. One is the investment appraisal of the acquisition of additional information on salinity, which is a really tricky thing: knowing the marginal benefit and marginal costs of getting more information on salinity. The second is the actual investment appraisal of the salinity management actions that happen down the track. The salinity mapping needs to be able to inform those investment options down the track, and section 2.2 probably needs to draw that distinction out.

Click on image for a larger version of figure 2

The most important question for me is: why are we mapping salinity? I think 90 per cent of the reason is in the first point there [on the slide]. We are trying to inform public and private investment decisions, but mostly the public investment decisions. I don't know that we are really going to be able to put too much useful information down for private farmers, who know very well how their farm works. They are already using all that information to manage salinity at the moment; I would be surprised if our mapping was really going to change that. I don't really know for sure, but that would be my impression.

The public investment decisions about where we spend limited money are obviously very important, and that is the guts of why we are doing a project like this. That probably needs to be articulated up front in the report, if it is.

Other reasons I can think of why you would map salinity are to advise industry and government on salinity's economic importance relative to other problems we are facing – that can be useful – to target regulatory responses, as in Queensland, where you do tree clearing, and maybe lastly but probably less significantly in reality, just to build knowledge about how Australian landscapes function.

I think most of what we are doing is to inform public investment decisions.

So how are those investment decisions actually made? I want to briefly show you some formal economic appraisal techniques, using the information that comes out of the salinity maps, that I have actually applied in salinity.

Click on image for a larger version of figure 3

First and foremost, most economists will apply benefit cost analysis. We need to know all the cost items, all the benefit items, what happens in each year or time period over the life of the project. We will set a discount rate and we will determine the net present value or benefit cost ratio and an internal rate of return. This is how we work out whether something is a good investment, and there are specific sorts of information we will need from the salinity mapping. There is a lot of philosophical debate about hazard and risk; as an economist I don't really care too much about that. I want to know what is happening to the resources we are trying to protect, in time periods following some sort of intervention. That will allow me to select the best-performing project. The problem with this approach is that it requires dollars on both sides of the equation, which we can't always supply.

Click on image for a larger version of figure 4

Cost effectiveness analysis is another formal approach to doing these investment appraisals, where we would, say, try with a budget to minimise the amount of salt that is going to end up in a river. So we would work it in a similar way but we are not trying to put dollars on the benefit item; we just have one attribute, which is tonnes of salt, and we look at the most cost effective way of doing that. It still involves discount rates and costing.

Click on image for a larger version of figure 5

Cost utility analysis is something only just coming over the horizon, so this is all very new. This is where we have multiple attributes of benefit arising from an investment in salinity, such as the biodiversity through to the social end as well. We construct a utility function and try and get the most cost effective way of achieving that, via some sort of multiple criteria analysis.

 

Figure 6	Figure 7
Click on images for larger versions

This is also a formalised investment appraisal technique which lets us pinpoint, via a set of multiple social, economic and environmental criteria, where we might allocate scarce resources. I have here [on the first slide] a basic example of multiple criteria analysis, which may not be so well known. We have a set of salinity management projects down the left-hand side, a set of criteria at the top, which might be Economic, Biodiversity and others that are non-monetary. We weight them and we apply algorithms to work out which one has got the best performance. That is a basic performance to MCA and is actually what NHT and NAP groups are most commonly using at the moment to evaluate which projects are funded, but in a much less formal way.

So those are a set of formal economic appraisal techniques for which your salinity maps are actually going to get used to make decisions.

Click on image for a larger version of figure 8

I might just highlight very quickly some other work we are doing here, the Strategic Landscape Investment Model, which is another way that the salinity mapping will be working. But again what is critical here is not your maps of hazard and risk but how that patch of dirt responds over time to some kind of treatment – the marginal benefit we get out of it. It allows us in this model to work out the optimal investment pattern across New South Wales, at this stage, where we would spend money. So it is more than just a map of risk or hazard.

I will go very quickly through some benefit cost analyses I have been involved in of salinity management in Australia. The economics of salinity is very grim, especially from a public point of view. We typically conclude to walk away. But the government is committed to spending some $1.4 billion under the NAP and will spend it; we still have to find the most efficient way of doing it.

Click on image for a larger version of figure 9

One study we did on the Lower Eyre Peninsula found that the net present value shortfall – the loss in money – was $173 million.

Click on image for a larger version of figure 10

That was going to save around 4 per cent of the catchment from being salt affected over a 20-year period. There is absolutely no way that would be seen as economic. Whatever the non-market or non-monetary benefits were of saving that 4 per cent of salt land over that period, they had to be worth $173 million, or roughly $2,000 per household per year. You could flip that benefit cost ratio over by making the population of Port Lincoln at about 1.5 million instead of 10,000 – to give you an example of how those things work.

Not just our study but Agriculture WA had a look at this. They looked at nine salinity revegetation projects, only two of which had benefit cost ratios above 1; the rest had very low benefit cost ratios of 0.15 to 0.45. The reason salinity works this way is long time lags, we are getting told by the hydrogeologists that you have to stick trees over a lot of the catchment to do something, it is making it look uneconomic. All this is relevant to the value of information on salinity we are going to get as a nation as well.

Analysing the economics of salinity at the national scale is something I was involved in through the National Land and Water Resource Audit. Here we mapped agricultural profits across the whole country, and then put maps of salinity that were produced in the audit against those to look at the yield decline and the consequent drops in profit.

Click on image for a larger version of figure 11

These were the main results: profit was only going to really decrease across the country by about 1 to 2 per cent because of salinity problems. That used the audit's salinity mapping, which required a lot of massaging to put into an economic interpretation. Again the problem there is that they sort of mapped risk and hazard, they applied different techniques in the different states so we didn't have something terribly comparable, and as an economic team we had to put all that biophysical data together in a way it could be used. We would have much preferred a best estimate or a best shot at mapping salinity extent – where the stuff is, where it is causing a problem, and what is going to happen in the next 20 and 50 years would have been a much more useful product.

This is another thing, that if the biophysical science community can't get those judgments right, or is unwilling to make them, they will be made either implicitly or explicitly by people less qualified to do so.

Click on image for a larger version of figure 12

All those numbers – if you can digest that, you're doing well. This is basically telling us that what we do find from the data we have got so far is that the problem of salinity is much less, in terms of the increase of profit of agriculture, than, say, for sodicity or acidity, other soil problems that are out there. It is a very minor one, we think, on the field.

That is a very hard-nosed look at it with the best available data we have got, which I reckon is still not very good.

 

Figure 13	Figure 14
Click on images for larger versions

Just to wrap up: some quick key messages of an economist's thinking about this problem of how you map salinity and how we advise governments. I am not sure that vague notions of risk and hazard are helpful. If you can qualify those notions of risk and hazard they will be good, but the maps I have had so far to use at a national scale were unqualified maps of risk and hazard, differently defined in different locations, we didn't know where it was – we just wanted to know where the actual salty ground was, what it was doing to crop yields and roads and houses, and then we could get a bit of a grapple on the benefits associated with managing it.

If risk and hazard can be defined and applicable, that is good. I should also say, though, that there is another application in the regulatory impact assessment stuff. For example, in Queensland, where do we prevent tree clearing as a priority? Then you do need to know about risk, the probability of hazard occurrence into the future. You want to try and keep trees where there is a high risk.

You need to relate the salinity severity to the actual impact, which I think is what I have just said. We need to know about yield loss and infrastructure damage.

The future trends are really critical for investment decisions in salinity. We know this stuff happens gradually, over long periods of time. The use of discount rates in economics is a real problem. Governments, as they make decisions, do value things in different time periods at different rates. You can disagree with that notion, but it is what is used and it does have some relevance when you are considering salinity investments versus health care investments versus building roads. So we need to know what is happening to this salinity problem over time, not just a snapshot.

And we need to know how it responds to our treatment. There is an issue that inconsistent methods in different locations will mean inconsistent economic analyses. If we do different mapping in all the states and territories of Australia, as was done in the audit, we are unable then to tell you about the relative economic magnitude of the problem in Queensland versus Victoria, or if it is across different regions the same will apply – we are unable to make those interregional comparisons very well unless a team of economists somehow standardise the data.

As I said, there is a critical need for data on the salinity response to treatment. That is what we are really trying to get at, because that is what we are investing in. Either the biophysical scientists provide these estimates as best you can, so you push yourselves to do what you feel uncomfortable doing, or we are going to do it, or it is not going to get used at all and the investment decisions made by government at the end will be worse.

The last point is that decisions will always be made with imperfect information. We would prefer 'best available' over nothing.

That is a quick take on how, as economists, we are seeing the problem of salinity mapping and the information we need.