Earlier this week, the Vanguard reported on the work that Dan Carson did on behalf of the city’s Finance and Budget commission. This argued that the 2013 Nichols Consultant report had overstated cost increases to pavement repairs needed for roads, sidewalks and bike paths on the order of $50 million, dropping the projected costs from $154 million down to $103 million.
Mr. Carson added, “Because the city does not have the financial resources in place to catch up immediately on the existing backlog of work, or to prevent any future backlogs from occurring, the true cost of the work is certain to exceed even my lower ‘budget needs’ estimate. Costs will go up as projects are delayed for lack of funding.”
Nevertheless, he concluded that “the Nichols report estimates of how costs would escalate because of constraints on available funding are, like their estimate of ‘budget needs,’ significantly overstated.”
Earlier this week, we accepted Mr. Carson’s premise and instead argued that, even at $100 million, the costs were significant.
In this column, we will unpack Mr. Carson’s analysis a bit more. He writes, “The Nichols model assumed that the costs of such projects would grow at an annual compounded rate of 8 percent.”
This is based on a weighted average of two specific inflators – one from SACOG and one from CalTrans. Mr. Carson argued that “there are problems with both inflation factors that Nichols used in its calculations and that the report overestimated the future growth in city pavement rehabilitation costs.”
From 1999 to 2012, the Nichols report cites the 15 percent annual growth in the California Asphalt Price Index as a partial basis for the assumed 8 percent annual growth assumption, and Mr. Carson notes that, by his calculation, the increase was actually 23 percent.
However, Dan Carson argues there are problems with relying on the Caltrans index. He writes, “Nichols assumed that city project costs would cumulatively go up about 8 percent in 2013 and 8 percent again in 2014. However, the Caltrans index has not tracked upward in line with those projections. Instead, it has cumulatively dropped 28 percent during the past two years.”
Secondly, Mr. Carson argues that the Caltrans index does not directly measure asphalt prices but rather, “It is calculated using the median of posted crude oil prices.” Mr. Carson examined asphalt data that shows that “asphalt costs have grown only modestly since January 2013, the start of the Nichols forecast period, and are now dropping.”
He finds, since 2000, that changes in oil prices “have typically resulted in corresponding changes in asphalt prices about three-fourths as great.”
Third, Mr. Carson argues the “problem with reliance on the Caltrans index is that using long-term historical crude oil price data alone to project future costs of the city’s pavement rehabilitation program does not take into account major changes that have occurred in the structure of the world economy since the 2008 recession.”
He notes that, since 2008, “crude oil prices have generally been on a downward slide, and plunged recently.”
Dan Carson writes, “The projections of an international forecast firm I have reviewed suggest the trend in crude oil prices through the year 2020 could be even more moderate – about 1.8 percent annually.” He adds, “One Southern California oil driller told me that, when his firm develops its financial plans for long-term projects, it assumes that crude oil prices will remain flat for the foreseeable future because of the instability and uncertainty of the market.”
Most notably he argues, “In contrast with this industry practice, the 8 percent annual compounded growth rate used by Nichols for its estimates implicitly assumes that the price of crude oil would grow to about $423 per barrel by 2032, the final year of the pavement rehabilitation program. The price of crude oil is currently about $69 per barrel, by one widely used industry benchmark.”
A fourth problem is that “the index is based unilaterally on crude oil and asphalt costs even though city pavement rehabilitation projects include significant costs for other types of building materials and labor. Construction project and labor costs have grown much more moderately than the 8 percent compounded inflation rate the Nichols report assumed in its projections.”
Based on these projections, Dan Carson updated the Nichols model with more moderate inflation assumptions. He writes, “I assumed that pavement rehabilitation costs would initially grow by 1.5 percent but by the end of the 20-year life of the program would escalate to 3.2 percent annually.”
He writes, “The Nichols report implicitly assumed, for example, that the cost of applying slurry seal to prevent road damage would grow from $4 per square yard in 2013 to $17.26 per square yard by 2032. My more modest inflation assumptions suggest that the $4 per square yard cost would grow to $6.20 per square yard in 2032 – a 55 per cent increase in costs rather than the 331 percent increase in costs assumed in the Nichols estimates.”
Using these numbers, he lowered the 20-year price tag by $10 million. Applying a similar approach for each of the components of the project, he reduced the overall cost for the roads portion by about $48 million.
When Dan Carson and I met last week, he was very explicit that he wanted to create a realistic model that could fairly accurately estimate costs. He was less interested in the policy conclusions that we might take from such an exercise.
Through Mr. Carson’s work we can see a few things. First the 1999 to 2008 period saw a very rapid increase in the cost of oil. After about a 15-year period of time where oil prices were fairly flat, the cost of oil would increase from $20 per barrel upwards of $100 per barrel.
Clearly, 2008 changed those dynamics, and the cost of oil has been relatively flat since the oil market rebounded from the crash. In the last six months, the price of oil has fallen dramatically, but most prognosticators believe that is temporary and that prices will rebound dramatically in 2015.
The problem I think we have in trying to reach a 20-year estimate is illustrated in what has happened in the last 20 years, since 1995. Would we have been able to predict that in the summer of 1998, for example, you could purchase gas at the low rate of $.79 per gallon only to see it explode for the first time well past $2, then $3, then $4 per gallon? Then the oil prices collapsed following the recession in 2008, and slowly rebounded back up to $4 a gallon, sometimes higher.
Would we have been able to, in 1995, foresee the 2001 attack on the World Trade Center that led to war in the Middle East? The 2008 Great Recession?
I bring this up to illustrate that there are fundamental flaws using the 1999 to 2012 period to project the cost of asphalt because the cost of oil dramatically went up. But on the other hand, Mr. Carson’s assumption that the changes in the structure of the world economy since the 2008 will continue, as the economy continues to rebound, is probably just as problematic.
I think we need to remember that we are making 20-year projections here. For illustrative purposes only, CalPERS bases its actuarial projections on a 30-year running average where it assumes an annual rate of return. Given the vitality of year to year returns, CalPERS uses the average rate of increase rather than worrying about year to year wild fluctuations.
Based on that principle, the notion that the Nichols report assumed an 8 percent increase in costs in 2013 and 2014, but the reality saw a 28 percent decrease, is less troubling. Over a 20-year period, one could reasonably believe that we would see those discrepancies cancelled out.
One advantage that CalPERS has is the ability to adjust on the fly. So if they see projected future earnings decline, they can reduce their assumptions which will force governments to increase inputs.
In the case of road repairs, the city has to plan its costs in advance. If the city were to pass a parcel tax, it needs to base the rate of the parcel tax and any bonding on reasonable current projections. While we may be able to adjust year to year spending from general fund sources, we get only one real shot at setting the rate for the parcel tax and bonding off it.
So given that, if, from 1999 to 2012, costs increased 23 percent per year, and the Nichols report lowers that to about 8 percent, is that an unreasonable assumption to calculate the future costs? This is a case where a conservative assumption means you guess a little high on the costs and the worst case scenario is that you end up with some additional funding for other infrastructure needs — all of which the city is not in good shape on.
At the end of the day, whether the true cost is $100 million or $150 million or more realistically somewhere in between, those are big numbers that we do not have.
Mr. Carson also notes that currently the city has carved out enough general fund money, about $3 million a year, to go for roads. With inflation factors, he projects that to mean the city would have more than $96 million available between now and 2013 for road rehabilitation, which aligns with his budget needs for those projects – even though he earlier acknowledged that costs would undoubtedly increase from that projection.
However, that kind of projection is what got us into trouble in the first place. When times got tough in 2008, the city balanced its budget by cutting back on non-labor costs, putting road repairs and other infrastructure into unmet needs categories.
In 20 years, we are likely to see several economic downturns and while they may not be as severe as 2008, why risk it? By relying on general fund monies, we take huge risks here that, in lean times, we won’t cut back on spending for roads and infrastructure. However, by creating a revenue stream devoted to infrastructure we avoid that risk.
The bottom line, the Nichols report probably represents a high end of the estimated costs, but smart planning should likely seek to use a conservative assumption by assuming higher costs and creating the revenue stream that can go to other infrastructure needs if the assumptions turn out to be too high.
Furthermore, whether it’s $100 million or $150 million, it is a large sum of money that will only increase if we fail to act.
—David M. Greenwald reporting