So why does it matter? Repairers are more and more up to their eyeballs with pandemic issues and adding the layer of knowing how to set posted rates is just too much to bear.
The rate is the basis of most of the business’ gross sales. It is the most important segment that needs regular attention. It is hard to understand why this is such a problem area in collision repair businesses, until we factor in all of the other things you have to do.
Many managers or owners are not highly educated in the ways of accounting and cost analysis. Therefore, it becomes a low priority as long as there’s money in the bank to ‘cover the payroll come Friday.’
Unfortunately, this will likely put the shop in the position to have to take out loans for upgrades, new equipment, certification feeds or technician training. When that debt stacks up, it becomes a problem to keep up with and is usually the point at which a realisation occurs: “We’re not making enough money to improve.” Then the decision whether or not to grow becomes so much harder and a source of great stress. You have to know everything about your numbers.
A recent insurance labour rate investigation showed an inability to keep pace with inflation and innovation and costs while playing the same games with suppressed labour rates.
Taking an in-depth view into the labour rate issue in the US, the issues of the current situation was considered by looking at the current actual levels of labour rates in a market – not controlled by auto insurers – and using widely accepted methods of economic valuation to show why they would exceed the level of auto mechanical labour rates.
There are also two opposite arguments: the auto insurers’ ‘market power interpretation’ of auto insurers’ control over their market with a discussion of ‘monopsony power’ (a market situation with only one buyer) and how the widely disparate rates of inflation of motor vehicle insurance premiums and collision repair costs support a ‘market power’ interpretation of this situation.
We will have to test the last two views based on statistical analysis of the labour rate.
The results of statistical analysis strongly suggest that the geographical patterns of labour rates being paid on collision repair claims by auto insurers does not conform or align with the pattern of costs (as reflected in the wages paid for auto collision or mechanical repair services) across US counties, based on variations in their population and shop densities.
Though wage costs consistently rise with these density measurements, labour rates do not – and in some cases they actually decline. The findings so derived strongly support a market power interpretation of this pattern over auto insurers’ false efficiency claims.
On becoming aware of auto insurers’ control over the collision repair industry, I was asked to help with an auto glass case in 1993. The pattern was similar to one I had seen in the healthcare industry. By 1994, I had extended my understanding enough to suggest a ‘Silver Bullet Test’ of the ‘efficiency versus market power’ arguments – if sufficient labour rate data were available for such a test.
The National AutoBody Research is collecting claims data from independent collision repair shops across the U.S., so I contacted them and was sent a database of 7 068 auto insurance claims from a diverse array of collision repair shops across 39 states, covering approximately 165 US counties. The statistical analysis that I then conducted compared the body labour rate paid in these claims with two different county-based measures of regional destiny – population and shops – both per square mile. County data were used because US census data on wage costs are also available on that basis.
The basic question
Does the geographical pattern of labour rates consist more closely with an ‘efficiency argument’ or with a ‘market power interpretation’? For an efficiency argument to prevail, labour rates should track wage costs. For a market power interpretation, labour rates should be less positively (or even negatively) related to population or shop density across different countries.
The key issue has to do with wage costs which rise with population density (wages tend to be higher in urban than in rural areas), which means that labour rates, under an efficiency argument should rise in accord with population or shop density to reflect these rising costs. A market power interpretation of these industry patterns, alternatively, would suggest that labour rates do not rise as much as costs do, or might even decline, with increasing population and shop density.
The reason is that, with more people and shops in a given county, there is more opportunity for auto insurers to play these shops off against each other in an economical, anticompetitive power game. Consequently, to the extent that market power abuse is occurring, labour rates should not respond to and thus reflect these rising costs as much or at all as population or shop density increases.
Looking at wage cost differentials across US counties, I then looked at the relation of labour rates to costs across those different counties for three different groups of claims. The first group was the entire data set, first with and then without the claims that were ‘customer paid’ (and thus non-insurance related). The second group focused on those auto insurance companies for which I had enough claims and counties to perform a statistically significant test of these data. The third group was to look at the states for which I had enough claims and counties to analyse them independently.
The measure of costs
Starting with the “Average Weekly Wage’ for three categories of services – (1) all services, (2) automotive mechanical and electrical repair and (3) automotive body and interior repair in all US counties. Interestingly the average wage in both all and auto mechanical repair are strongly and positively correlated with both population and shop density across all US counties, while average wages in auto body repair have only a slight positive correlation with these two measures of density.
This suggests that auto body repair wages do not track cost differentials as well as do services in general or wages in auto mechanical repair. One might conclude from this finding that there is something abnormal going on in the collision repair industry that is keeping these wage costs from rising in accord with the wage costs for auto mechanical repair or for services in general.
Across the entire sample of 7 068 auto insurance claims, the average weekly wage (AWW) in both auto collision and auto mechanical repair by county is related to population and shop density in a strongly positive way, confirming that labour costs do rise significantly with population and shop density across counties.
In contrast body labour rate paid (BLRP) showed a weakly negative correlation with both density measures. Consequently, collision repair labour rates are not aligned with costs over the entire sample. This suggests strong support for a market power interpretation of these labour rate patterns over an efficiency argument at well over a 99.9% significance level.
A directly negative correlation of body labour rates paid with wage costs (AWW) in both auto collision and mechanical repair (ignoring density measures) also reinforces a market power interpretation, as it further implies a lack of alignment between labour rates and wage costs.
Group two – Individual Auto Insurance Companies
My first set of findings for auto insurance companies was for ‘all majors’ as a group of significant interstate auto insurers. Within this group of about 3 400 claims. Body labour rates paid were negatively correlated with both shop and population density, while the correlation of auto body repair and auto mechanical repair wages with both density measures was strongly and significantly positive. So, once again, within this selected group of 3 400 claims, labour rates do not align with costs, so this finding strongly supports a market power interpretation over an efficiency argument as well over a 99.9% level of significance. This result is also reinforced by a direct negative correlation of labour rates with wages in both types of auto repair across counties.
I looked at all individual auto insurers with more than 40 claims. Of the four insurers with 400 or claims, labour rates were either strongly negatively or weakly positively correlated with both density measures and differed from wage costs at mostly more than a 99.9% significance level. The direct correlation of labour rates with both wage cost measures, however, was negative only for two insurers and only weakly positive, (near to zero) for the two others. So, the results for these four auto insurers strongly support a market power of interpretation of their labour rate patterns.
Among six additional interstate insurers with between 100 and 400 claims in the dataset, only two had no strong difference between labour rates and population or shop density. The rest had significant differences between labour rates and density with significance levels ranging among 95%, 99% and 99.9%. So once again, even with these smaller sample sizes, most of the test results support a market power interpretation over the auto insurers’ efficiency argument.
Were there state-to-state variations in this data? I looked at the claim forms from specific states and the variability of wage costs across state counties for 15 of the 39 states for which there were 100 or more claims. The results for 10 of these states supported a market power interpretation of the labour rate patterns, mostly at 99.9% level of significance. The direct correlations between labour rates and either auto body repair or auto mechanical repair wages were also either negative or only weakly positive in nearly all of these states.
While acknowledging the inherent limitations of the available claims data, the important question addressed by the ‘Silver Bullet Test’ does have a definite answer. The regional distribution of labour rates in the 7 068 claims examined in the study yields a significant level of strong support for the auto collision repair industry’s ‘market power interpretation’ of the control over their industry exercised by auto insurers, and the results shown do not support the auto insurers ‘efficiency argument’ that has been used to justify these insurers’ control over the auto collision repair industry and their suppression of labour rates.
The geographical pattern of labour rates paid on these claims does not align with the increasing wage costs associated with higher population and shop density across counties, while such an alignment is required by any efficiency claim.
The apparent fact that (in most cases) labour rates either decline or increase very little with rising population and shop density, while wage costs increase significantly, amounts to evidence supporting an effective suppression of labour rates stemming from an abuse of market power in the collision repair industry resulting from auto insurers’ control over the process of auto collision damage appraisal repair.
This conclusion arises from the fact that higher population and shop density make it easier for auto insurers to play auto collision repair shops off against each other in an anticompetitive abuse of monopsony power through a price-profit squeeze against them, while low density areas make such anti-competitive ploys more difficult or impossible for auto insurers to implement effectively.
The final conclusion is this – auto insurers are, through an abuse of their market power, artificially suppressing labour rates below what are fair and reasonable in an anti-competitive manner.
Fred Jennings, Jr, of EconoLogistics (“Consultants to Business and Law”) can be contacted at firstname.lastname@example.org