I respectfully dissent. The majority concludes that the 1980 amendment to Government Code section 317201 (the 1980 amendment) merely clarifies, without substantive change, the existing test for industrial causation for service-connected disability retirement. (Ante, p. 578.) In reaching this result, the majority adopts the judicially created test for causation set out in DePuy v. Board of Retirement (1978) 87 Cal.App.3d 392, 398-399 [150 Cal.Rptr. 791, 12 A.L.R.4th 1150], a preamendment case. I disapprove of applying the DePuy test because it does not address the requirement, added by the 1980 amendment, that the employment “contributes substantially” to an incapacity in order to qualify for a service-connected disability retirement. Instead, DePuy espouses two standards, *582neither of which provides relevant guidelines for interpreting the 1980 amendment: a preamendment test for industrial causation, and an evidentiary test to determine if such causation has been proved. In my opinion, the Legislature intended a substantive change when it enacted the 1980 amendment and the majority has failed entirely to deal with the difficult problem of giving effect to that change.
I. The History and Application of the 1980 Amendment
Resolution of the meaning of the 1980 amendment first requires considering the chronology of events that led to the Legislature’s amendment of section 31720.
A. Rau
An applicant for disability retirement benefits is entitled to what is generally a higher pension rate if the disability is service connected. The standard for finding a “service-connected” disability is set by section 31720. Prior to the 1980 amendment at issue (eff. Jan. 1, 1981), that section provided in pertinent part: “Any member permanently incapacitated for the performance of duty shall be retired for disability regardless of age if, and only if: (a) His incapacity is a result of injury or disease arising out of and in the course of his employment. ...” (Italics added.)
In Rau v. Sacramento County Ret. Bd. (1966) 247 Cal.App.2d 234, 236 [55 Cal.Rptr. 296], one of the first cases to mention unamended section 31720, the court characterized the section as providing a “comprehensive legislative enactment for the determination of the rights of a county employee to be retired for disability.” In concluding there was sufficient evidence to support the board’s decision, Rau did not, however, discuss the portion of section 31720’s language requiring that the disability be a result of an injury or illness arising out of and in the course of employment.
B. Heaton
Not until 10 years later did the Court of Appeal for the first time directly consider the meaning of the applicable language of section 31720. (Heaton v. Marin County Employment Retirement Bd. (1978) 63 Cal.App.3d 421 [133 Cal.Rptr. 809].) In Heaton, a county employee claimed a service-connected disability pension based on a mental disability and asserted her employment was the principal cause of the disability. Despite the hearing officer’s finding that the employee’s disability was service connected, the board denied the application. The Court of Appeal affirmed the superior court conclusion that (1) the board’s rejection of the hearing officer’s find*583ing constituted an abuse of discretion, and (2) the employee was entitled to a service-connected disability retirement.
In Heaton, the board challenged the trial court’s interpretation of section 31720 as providing that even if a disability arose for a variety of reasons, the fact that a part of the disability was employment-caused provided a sufficient basis to award a service-connected disability pension. Instead, the board contended that section 31720 required “the employment be the sole or at least the substantial contributing cause of the disability before retirement disability may be awarded.” (63 Cal.App.3d at pp. 425-426, italics added.)
Finding this a question of “first impression,” the Heaton court rejected the board’s argument. It reasoned: “A simple reading of . . . section 31720 does not disclose the requirement which [the board] would propose.” (63 Cal.App.3d at p. 428.) The court emphasized that the section required only that the incapacity be “‘a result of injury or disease arising out of and in the course of his employment,’ not the result thereof.” (Ibid.) The court explained that the interpretation suggested by the board would require it to “rewrite the statute” and that the court’s function was not to legislate or to severely limit the statute’s then “present applicability.” (Id. at p. 429.) The board’s remedy, the court observed, was with the Legislature. (Id. at p. 431.) Thus, the Heaton court rejected the board’s attempt to set the quantitative measure of industrial causation at a substantial level, based on the express statutory language which at the time required only that the employment be a cause of the disability.
C. Subsequent Cases—DePuy
Various courts interpreted Heaton to hold that even an “inconsequential” or “infinitesimal” connection between the disability and the employment entitled an employee to a service-connected disability retirement. (See, e.g., Van Hook v. Board of Retirement (1983) 148 Cal.App.3d 714, 716, fn. 1 [196 Cal.Rptr. 186]; DePuy v. Board of Retirement, supra, 87 Cal.App.3d at p. 396.) These interpretations apparently followed from the Heaton court’s rejection of the board’s argument that where incapacity is due to mental disability, employment must be the sole or primary cause in order to prevent an “infinitesimal contribution” from justifying full compensation. (Heaton, supra, 63 Cal.App.3d at pp. 430-431.)
The appellate court in DePuy attempted to moderate the application of an “infinitesimal contribution” standard by requiring that while employment-related disability could be a very “small” part of the causal factors, it must nonetheless be “real and measurable.” (87 Cal.App.3d at p. 399.)
*584The DePuy court also set forth the applicable evidentiary standard: “There must be substantial evidence of some connection between the disability and the job.” (Ibid.) DePuy thus interpreted the preamendment version of section 31720 in light of Heaton, as well as enunciating the appropriate measure of proof. Its discussion thus addressed some of the more extreme interpretations of Heaton, but did not purport to reconsider the issue of “substantial contribution” raised by the board in the earlier case.
II. The Meaning of the 1980 Amendment
A. Legislative Intent
The legislative history indicates that following Heaton and its progeny, various counties urged a change in existing law, fearing that liberal interpretations of industrial causation would cause “the cost of county disability programs [to] grow to prohibitive levels.” (Assem. File Analysis, Sen. Bill No. 1076, June 5, 1980; see Lundak v. Board of Retirement (1983) 142 Cal.App.3d 1040, 1045 [191 Cal.Rptr. 446].) As the majority accurately traces, the legislative response in enacting the 1980 amendment was intended to disavow the implications of Heaton, but not to go so far as to require that the disability be principally caused by an employment-related factor. (Ante, p. 577.) The Assembly retreated from its originally drafted requirement that the incapacity be the principal result of the employment-caused injury and substituted the current language which added to the original statute the requirement that the “employment contributes substantially” to the incapacity. This final form apparently reflected a compromise between the bill’s supporters and labor representatives who strenuously opposed any change. Based on this legislative history, I agree with the majority that ‘“contributes substantially’ means more than ‘any’ and less than ‘[principal].’” (Ibid.)
However, I disagree with the majority’s holding that the 1980 amendment’s only purpose was to disapprove the “infinitesimal contribution” language in Heaton and not the entire body of case law construing that preamendment section. (Ante, p. 577.)2 Instead, I find the legislative history and the plain language of the amendment demonstrates that a new quantitative measure for the amount of connection or causation between employment and disability was fashioned.
First, the fact that the Legislature chose to adopt totally new language rather than defining previous language indicates that the amendment was *585not intended to simply reaffirm the pre-Heaton or Heaton standard or to encapsulate the DePuy approach. If the Legislature had intended that the amendment only eliminate the “infinitesimal” language of Heaton, it could have done so expressly and directly.
Instead, the Legislature added a completely new formulation for service-connected disability by requiring that the employment “contributes substantially” to a disability. Moreover, the fact that this precise wording was a compromise result after an initial attempt to require that employment be the “principal” cause of disability further demonstrates an intent to develop a new and stricter standard for industrial causation.
This conclusion is also supported by the “Comment” to the Third Assembly Amendments dated May 6, 1980, which states: “This bill is one of numerous measures in response to a recent court decision, Heaton vs. the Marin County Board of Retirement, which stated that a member is entitled to a service-connected disability if (1) he or she is permanently unable to perform his or her job, and (2) any part of the disability is job-connected.” (Original italics.) Thus, the expressed reason for the bill shows the Legislature’s response addressed the quantitative measure of industrial causation and was intended to reject the idea that any employment connection to the disability was sufficient.
As noted in Heaton, the court rejected the board’s suggestion that the employment “contributes substantially” to the disability in favor of a standard that “any” contribution would suffice. The Legislature soon thereafter amended the statute to explicitly provide for a substantial contribution requirement, one of the standards the board originally proposed in Heaton. It seems implausible to suggest, as the majority does, that the 1980 amendment signals a simple return to a pre-Heaton standard, requiring only that employment be “a cause” of the disability in order to qualify for a service-connected disability retirement. Even after the appellate court in DePuy rejected the post -Heaton interpretation of “infinitesimal contribution” and reinstated the standard of minimal contribution, the Legislature found further intervention was necessary to arrive at a workable standard.
The Legislature further indicated its belief that the 1980 amendment constituted a change in the law by providing that the substantial-contribution test “shall be applicable to all applicants for disability retirement on or after the effective date” of the amendment. If no change was intended, not only from Heaton's standard, but from the standard applied in DePuy, this additional provision would have been unnecessary.
Thus, the majority’s return to the DePuy test, requiring only that the employment’s contribution to the disability be “real and measurable,” *586does not properly take into account the intended effect of the 1980 amendment. Under DePuy, “any” connection other than an infinitesimal one will suffice so long as it is supported by sufficient evidence of a connection. The DePuy language relied upon by the majority does not address the changed quantitative measure of causation enacted by the Legislature. The majority has misapplied DePuy in interpreting the 1980 amendment and, in doing so, confuses the new language’s meaning and reduces its intended impact. Because the majority’s purported authority is of no assistance, and because the Legislature did not promulgate any other guidelines as to what is meant by “substantially” in the 1980 amendment, the proper course requires us to look to other sources.
B. Proposed Interpretations
Various tests to determine the meaning of the amendment’s “contributes substantially” language have been proposed. Petitioner Bowen urges that we follow the interpretation which was suggested in Lundak v. Board of Retirement, supra, 142 Cal.App.3d at page 1046, while the board suggests we look to common definitions for the word “substantial” or, alternatively, require a percent greater than 50 to determine industrial causation.
In Lundak, the court looked to the Restatement Second of Torts for assistance and analogized to the test for “legal cause”; namely, that negligent conduct is a “substantial factor” in bringing about harm. (142 Cal.App.3d at p. 1045.) In the context of tort law, the Lundak court stated the word “ ‘substantial’ is used to denote the fact that the defendant’s conduct has such an effect in producing the harm as to lead reasonable men to regard it as a cause, using that word in the popular sense, in which there always lurks the idea of responsibility, rather than in the so-called ‘philosophic sense,’ which includes every one of the great number of events without which any happening would not have occurred.” (Id. at pp. 1045-1046, quoting Rest.2d Torts, § 431, com. a.)
The Lundak test for “substantial” is inappropriate, in my view, to interpret the 1980 amendment. In Lundak, the court found that medical testimony regarding the contribution of the employee’s work-related accident to his disability clearly established that it was more than “inconsequential,” placing it somewhere between 10 and 60 percent. The court noted that the disability met the “real and measurable” test of DePuy and reasoned the tort definition of substantial would be consistent with the Legislature’s intent to restrict Heaton's liberal awards of service-connected disability benefits. Thus, the Lundak definition “would not include any contribution of employment to disability, no matter how small and remote.” (142 Cal.App.3d at p. 1046, original italics.) Yet, the Lundak *587court concluded: “We think reasonable people would regard [the employee’s] employment as a cause of his disability. . . . We hold, therefore, that . . . employment contributed substantially to his disability and that the terms set forth in subdivision (a) of Government Code section 31720, as amended, are satisfied.” (Ibid., italics added.)
I would conclude that Lundak does not represent an appropriate quantitative measure of industrial causation: all that is required is a cause or the “idea” of responsibility. This standard, like that in DePuy, will permit qualification for a service-connected disability retirement under a far more liberal measure than the Legislature intended when it amended section 31720. Lundak does not discuss the question of whether the 1980 amendment works a substantive change in the law. Instead, it seems to continue to employ the standards of DePuy, a result I believe the Legislature did not intend.
In concluding that the Legislature’s only intent was to reject Heaton and its progeny, permitting full disability pensions based on “infinitesimal” contributions, the majority ignores the basic rule of statutory construction which instructs that we should give effect to a statute according to the usual ordinary import of the language employed in framing it after looking to the legislative intent. (Moyer v. Workmen’s Comp. Appeals Bd. (1973) 10 Cal.3d 222, 230 [110 Cal.Rptr. 144, 514 P.2d 1224]; Rich v. State Board of Optometry (1965) 235 Cal.App.2d 591, 604 [45 Cal.Rptr. 512].) Thus, we must examine the language added by the 1980 amendment to determine what standard would accord with the effect sought by the Legislature.
The board advances commonly used definitions such as “main part of,” “significant,” and “preponderance” to elaborate on the meaning of the amendment’s requirement of “contributes substantially.” It alternatively suggests a minimum 50 percent industrial-causation test.
I do not believe the test of causation should require 50 percent causation from employment. The legislative history instructs that the word “principal” was replaced with “substantial,” indicating the quantitative measure was intended to be less than 50 percent or the primary cause. However, to contribute “substantially,” the amount attributable to employment must be more than a “small” but “real and measurable” amount of contribution. I believe that more is necessary than the 10 percent or less that DePuy and Lundak suggest.
Based on both the plain meaning of the statute and the legislative history, I would conclude employment must contribute more than 10 percent to a disability in order to qualify for a service-connected disability retire*588ment under the 1980 amendment. In my view, a test requiring that a greater percentage of the disability be employment-caused would reflect the intended meaning of the 1980 amendment while serving broadly and fairly to award service-connected disability pensions to qualified individuals.
Such an interpretation would further the purpose of the amendment by relieving the financial pressure on counties in awarding service-connected disability retirements. Legislative hearings prior to the enactment of the 1980 amendment indicated that the concern generated by the minimal contribution approach arose out of the fact that it can be argued that almost any injury or disease is service connected to some very small degree.
Statistics were presented at the legislative hearings for the previous 20 years which demonstrated a very rapid increase in service-connected disability retirements following the decision in Heaton. The cost of disability pensions in one county alone was projected as amounting to almost $1 million annually within the next five or six years. Although disability claims had risen in general over the years, the accelerated increase following Heaton threatened the system, and resulted in the move for legislative change.
A requirement that industrial causation amount to more than 10 percent would serve the goals of those both seeking and enacting the amendment and end inappropriate awards of service-connected disability retirement based on disabilities whose genesis had very little to do with the service itself. It would secure such awards for those who should be entitled to compensation because their employment had contributed substantially to their disability, thus preserving the financial integrity of the system for all those who fairly qualify for disability pension benefits.3
III. Conclusion
In sum, I disagree with the majority that the 1980 amendment did not substantively change the test for service-connected disability. In my view, the 1980 amendment imposed a new and higher standard for industrial causation, which clearly constitutes a substantive change. Such a change would, of course, necessitate considering the arguments regarding retroactive application, including possible limitations due to constitutionally vested pension rights. By concluding that the 1980 amendment signals only a return to the pre-Heaton standard of industrial causation, thereby avoiding the need to consider its retroactive effect, the majority evades its re*589sponsibility to interpret the statute accurately and to provide guidance in applying pension legislation in a fair manner and in accordance with the Legislature’s intent.
Respondent’s petition for a rehearing was denied October 30, 1986. Lucas, J., and Panelli, J., were of the opinion that the petition should be granted.
All further statutory references are to this code unless otherwise specified.
Of course, the entire relevant body of case law was post -Heaton, thus reflecting or in some manner responding to that case’s views.
Of course, nothing in this formulation would alter a person’s entitlement to regular pension or nonservice-connected disability benefits.