Turner v. State Ex Rel. Department of Transportation

*646 NAKAMOTO, J. In this personal injury action arising out of a motor vehicle accident, plaintiff Turner contended that defendant Oregon Department of Transportation (ODOT) had contributed to the accident by negligently failing to correct hazardous conditions at the highway intersection where he was injured. In a summary judgment motion, ODOT claimed immunity from liability under ORS 30.265(6)(c), a provision of the Oregon Tort Claims Act, ORS 30.260 to 30.300, that immunizes governmental entities from tort liability for “the performance of or the failure to exercise or perform a discretionary function or duty.” ODOT’s claim of immunity rested on the theory that, because it has a policy of seeking highway safety improvement funding from the legislature only for the most crash-prone sites in the state highway system, ODOT’s omission of the crash site from its appropriation requests amounted to a budget-driven, discretionary policy decision not to make improvements at the site. Plaintiff opposed the motion with evidence that ODOT employees knew that the intersection was dangerous but had neglected to make improvements by using ODOT’s other mechanisms for evaluating and funding low-cost highway safety improvements. Although ODOT prevailed in the trial court, the Court of Appeals determined that questions of fact precluded summary judgment in favor of ODOT. Turner v. Dept. of Transportation, 270 Or App 353, 365-68, 348 P3d 253 (2015). On review before this court, petitioner ODOT argues that, when a state agency uses a global process for setting priorities and allocating limited resources, discretionary-function immunity attaches and the agency need not engage in further, particularized decision-making. But whether or not we agree with ODOT’s proposition in theory, ODOT’s immunity argument rests on the premise that its crash-based ranking process for appropriation requests was global—that is, a comprehensive method for determining which safety improvements to fund—and so its failure to improve the intersection at the crash site may be ascribed to policy choices ODOT made in its appropriation requests. The record on summary judgment does not bear out that premise as a matter of undisputed fact. It follows that ODOT’s *647 employment of the ranking process cannot resolve the issue of ODOT’s immunity under ORS 30.265(6)(c) as a matter of law. Therefore, the trial court erred in granting summary judgment for ODOT on that ground, and we affirm the decision of the Court of Appeals. I. FACTS AND PROCEDURAL HISTORY We present and consider the facts in the light most favorable to the nonmoving parties. Towe v. Sacagawea, 357 Or 74, 77 n 2, 347 P3d 766 (2015). Collins Street in the City of Depoe Bay runs into State Highway 101. In 2008, when the accident in question occurred, motorists driving west on Collins Street could turn onto Highway 101 in either direction. Entering the highway was difficult, and particularly so for those turning left, to the south. The topography of the area and vehicles parked on the highway limited the line of sight needed to safely enter the highway. In 2008, as defendant Colip was turning left from Collins Street onto Highway 101, her car collided with plaintiff, who was on his motorcycle riding north on Highway 101. Plaintiff was severely injured in the accident. Three governmental entities had the ability to change the conditions at the intersection before the accident occurred: ODOT, which owns Highway 101; Lincoln County, which owns the relevant part of Collins Street; and the city, where the intersection is located. ODOT, the county, and the city had been aware of the safety problem at the intersection since at least the mid-1990s. For example, a 1995 ODOT memorandum concerning the impacts that a proposed real estate development in the area would have on roadways stated that there was a significant accident history at the intersection: “The 1992 accident rate for an urban, primary system, non-freeway is 3.69. The actual accident rate for this area, from 200 feet north of Clarke to 100 feet south of Bay Street, is 4.61. As you can see, this is above the state average. * * * “My interpretation of the accident summary database shows a significant accident history in the Depoe Bay area. Bay Street and Collins Street intersection areas seem to have a significantly higher amount of accidents than other intersections in the downtown area. * * * *648 * * * * “An interim solution, regardless of the development size or staging, to increase safety on the highway, is to eliminate egress from Collins Street. This would eliminate the visibility restrictions to the south currently experienced by motorist [s] entering the highway. Region 2 Traffic Operations, Lincoln County, and the City of Depoe Bay Engineers should review proper signing and delineation/ barricade techniques restricting the westbound motorists on Collins Street.” That same year, Owings, the city’s superintendent for streets and public works, wrote to the city council and ODOT, warning them that turning south at the intersection was “very dangerous.” Later, the concerns addressed in the ODOT memorandum and Owings’s letter were relayed to county supervisors, including the county public works director. And, ODOT, the county, and the city had even discussed, at length, possible solutions for the Collins Street intersection. Those possibilities included restricting or reconfiguring parking on Highway 101 at the intersection, prohibiting left-hand turns from Collins Street onto Highway 101, and closing access to the highway from Collins Street altogether. All three entities had also participated in the city’s development of a transportation plan (2000), and a subsequent “refinement” plan (2005), both of which included proposals for improvements aimed at the line-of-sight problem at the intersection. Until the 2008 accident that is at the center of this case, however, none of the proposed improvements had been specifically planned, funded, or implemented by ODOT or the other governments. If ODOT had decided to undertake one or more of the suggested improvements, the project would have needed funding. The largest amount of funding that ODOT had for highway safety improvement projects derived from a multi-step process that ODOT used for seeking and obtaining appropriations from the legislature for transportation projects of various sorts, including highway safety improvements. *649 Kargel, the traffic engineer for ODOT’s Region 2 that included Lincoln County, explained in a declaration that, through ODOT’s Statewide Transportation Improvement Program (STIP), ODOT selected highway safety projects to be included on a list of projects for ODOT’s biennial budget request. ODOT did that by using a computerized crash-history list, the Safety Priority Index System (SPIS), “and according to a cost/benefit analysis of improvements to high accident sites.” Sites that fell in the top five percent of the SPIS or that were highly rated “based on a cost/benefit analysis” were included in a list of potential highway safety construction projects to be included in the STIP Safety Budget submitted to the legislature as part of ODOT’s biennial funding request. ODOT did not request any funding to address the line-of-site problem at the Collins Street intersection through the STIP process. ODOT also had other sources of funding to enhance safety. In her deposition, Kargel added that small amounts of money from an ODOT maintenance manager’s maintenance budget were available. She acknowledged that the maintenance manager for the Depoe Bay area could have used money to remove parking spaces at the intersection of Collins Street and Highway 101, for example. And, Kargel explained, if an intersection was not on the SPIS list, ODOT could still make a decision to fund low-cost physical improvements at that intersection, such as by removing or changing the parking available near the intersection. If a safety improvement project was not selected for funding through the STIP process, an alternative Highway Safety Engineering Quick Fix program was available, beginning in 2007. The Quick Fix program had “a dedicated bucket of safety funds” that was specifically designed for highway safety problems that are “best addressed by low-cost engineering countermeasures without going through the formal STIP project development process.” ODOT’s regional offices were required to administer the Quick Fix program. The record contains no evidence that ODOT conducted a cost/benefit analysis for improvements at the Collins Street intersection or analyzed whether a low-cost solution for improving the intersection was suitable under the Quick Fix program. *650 In his personal injury action against Colip, ODOT, the city, and the county, plaintiff alleged that the accident resulted from Colip’s negligence in failing to properly look out for and yield the right of way to oncoming traffic and from the three government entities’ negligence in failing to take steps to correct or warn of the unsafe conditions at the intersection. Colip filed cross-claims for contribution against the three government entities, reiterating plaintiffs allegations of negligence and adding allegations that were somewhat broader. Generally, plaintiffs and Colip’s allegations of negligence with respect to ODOT focused on ODOT’s failure to prohibit diagonal parking along the highway at the Collins Street intersection, restrict left turns onto the highway from Collins Street, post signs on Highway 101 warning drivers about the intersection, close the intersection to traffic, or take other measures to increase visibility at the intersection or to otherwise correct or mitigate the sight-distance problem. In their answers to plaintiffs complaint and Colip’s cross-claims, the three government entities raised discretionary-function immunity, ORS 30.265(6)(c), and the applicable statute of limitations, ORS 30.275(9), as affirmative defenses, and they eventually sought summary judgment against plaintiff on those grounds. Initially, the trial court granted their motions on the statute of limitations ground, without addressing discretionary-function immunity. Later, after the trial court declined to hold that the statute of limitations also barred Colip’s cross-claims, the three government entities moved for summary judgment on those cross-claims on the basis of discretionary-function immunity. Each government entity asserted a separate theory as to why its own inaction with respect to the Collins Street intersection amounted to a discretionary policy decision that was immune from liability under ORS 30.265(6)(c), and each submitted evidence in support of its theory. Ultimately, the trial court granted the motions for summary judgment on Colip’s cross-claims. At the close of the hearing, the court explained that, in its view, all three entities had been engaged in an ongoing planning process concerning the problems at the intersection and that their *651 efforts in that process brought their actions (or inactions) within the protection of the discretionary-immunity statute. Based on its summary judgment rulings, the trial court issued a limited judgment for the three government entities, effectively dismissing them from the case. Plaintiff and Colip appealed from that limited judgment, arguing that the trial court had wrongly decided both the statute of limitations and discretionary-function immunity issues. The Court of Appeals affirmed in part and reversed in part. Turner, 270 Or App at 372. For reasons that are irrelevant to our review, it concluded that the trial court had erred in granting summary judgment for the government entities based on the statute of limitations. Id. at 362-63. And, although the court concluded that the county had shown that there were no material issues of fact regarding its immunity from liability under the discretionary-function immunity provision, it held that the applicability of that provision could not be resolved as a matter of law with respect to ODOT and the city. Id. at 363-72. Of particular relevance here, the Court of Appeals rejected ODOT’s contention, which we describe more fully below, that the process that ODOT had employed to prioritize and select highway construction projects for its appropriation request immunized it from liability as a matter of law for any and all of the omissions that plaintiff and Colip had identified as negligence. The Court of Appeals concluded that that theory was unavailing when the record on summary judgment “d [id] not show that all of the Highway 101 modifications in question were considered and rejected in the [crash-history ranking] process or that other available processes were used to decide to not make those changes.” Id. at 367. 1 *652 Of the parties that had reason to be dissatisfied with the Court of Appeals decision, only ODOT sought review. This court allowed review to address ODOT’s contention that the Court of Appeals had erroneously limited the discretionary-function immunity to which state agencies are entitled under ORS 30.265(6)(c). II. DISCUSSION A. Discretionary Immunity Before we turn to the particulars of ODOT’s claim of immunity, we briefly describe the relevant legal landscape. Public bodies in Oregon are liable for the torts of their employees and agents, with certain exceptions. ORS 30.265(1). One exception is the so-called “discretionary function” exception, set out at ORS 30.265(6)(c). It provides: “Every public body and its officers, employees and agents acting within the scope of their employment or duties * * * are immune from liability for: «⅜? ⅜ ⅝ ⅜ ⅝ “(c) Any claim based upon the performance of or the failure to exercise or perform a discretionary function or duty, whether or not the discretion is abused.” What constitutes a “discretionary function or duty” for purposes of ORS 30.265(6)(c) has been the subject of considerable discussion in this court. In a nutshell, governmental conduct amounts to performance of a “discretionary function or duty” if it “is the result of a choice among competing policy considerations, made at the appropriate level of government.” Garrison v. Deschutes County, 334 Or 264, 273, 48 P3d 807 (2002). It is important to understand that not all decisions by governmental actors involve such policy choices. Discretionary-function immunity does not extend to “routine decisions made by employees in the course of their day-to-day activities, even though the decision involves a choice among two or more courses of action.” Lowrimore v. Dimmitt, 310 Or 291, 296, 797 P2d 1027 (1990). As this court explained in McBride v. Magnuson, 282 Or 433, 437, 578 P2d 1259 (1978), *653 “insofar as an official action involves both the determination of facts and simple cause-and-effect relationships and also the assessment of costs and benefits, the evaluation of relative effectiveness and risks, and a choice among competing goals and priorities, an official has ‘discretion’ to the extent that he has been delegated responsibility for the latter kind of value judgment.” The accepted rationale for providing discretionary-function immunity—separation of powers of coordinate branches of government, see Stevenson v. State of Oregon, 290 Or 3, 10, 619 P2d 247 (1980)—explains the particular focus on governmental policy choices and also provides insight into the immunity inquiry. A governmental actor performs discretionary functions and duties when exercising delegated responsibility for making decisions committed to the authority of that particular branch of government that are based on assessments of policy factors, such as the social, political, financial, or economic effects of implementing a particular plan or of taking no action. As this court has explained, “[w]hen a governmental body by its officers and employees makes [a policy] decision, the courts should not, without clear authorization, decide whether the proper policy has been adopted or whether a given course of action will be effective in furthering that policy.” Id. B. ODOT’s Summary Judgment Motion In ODOT’s motion for summary judgment against Colip, its discretionary-function immunity claim centered on its use of crash-history rankings to select safety improvement projects for funding requests from the legislature. 2 ODOT explained that, as a matter of policy, it has chosen to allocate its limited highway safety improvement funds to sites selected under the STIP. ODOT argued that its use of SPIS rankings and the STIP process to select projects for *654 funding means that ODOT’s nonaction with respect to a site that does not qualify under the SPIS metric is the result of a number of high-level policy choices: (1) a decision to use the agency’s limited budget on selected projects rather than spreading the money evenly over the entire highway system; (2) a decision to select projects according to relative dangerousness of the site; and (3) a decision to assess relative dangerousness based on crash statistics indexed in the SPIS. Taken together, ODOT asserted, those decisions constitute a global policy decision not to fund improvements at sites that are not within the top five percent of the SPIS or rated highly based on a cost-benefit analysis—a policy decision that falls within the protection of the discretionary-immunity statute. That policy decision implicated the Collins Street intersection, ODOT added, because it did not rank in the top five percent of the SPIS or qualify under the cost/benefit alternative at the relevant time: Even if ODOT was aware of the hazard at the intersection and wished to take steps to mitigate it, ODOT argued, the SPIS-to-STIP budgeting process meant that no mitigation project would be funded at the site. When ODOT moved for summary judgment in the trial court, it supported its motions with a single declaration, Kargel’s declaration. In it, Kargel described the STIP and SPIS processes in the following terms: “Highway safety construction projects are selected according to high-level ODOT budgetary policy using a computerized safety ranking process known as Safety Priority Index System (SPIS), and according to a cost/ benefit analysis of improvements to high accident sites. “ODOT prioritizes safety construction funds primarily on crash history as reflected in the SPIS safety statistics and the projected safety benefit that a project will have on that crash history. Specifically, it is ODOT policy to include the worst 5 percent SPIS-rated accident sites, as well as other high accident rated sites based on a cost/ benefit analysis, in a list of potential highway safety construction improvement projects in the Statewide Transportation Improvement Program (STIP) Safety Budget. “STIP Safety Budget funding requests for highway safety construction projects selected in this manner *655 according to ODOT policy are included in the State Highway Construction Plan (Plan) using the list of projects from the STIP safety budget. The State Highway Construction Plan is then submitted for funding to the Oregon Legislature with the Governor’s proposed biennial budget pursuant to ORS 184.658.” Kargel’s declaration also indicated that the Collins Street intersection had not ranked in the top five percent of the SPIS in 2008 and that the intersection had “not been considered a high-accident site before 2008”: “The crash rates for the five years before the 2005 Refinement Plan study was published show that Depoe Bay rates were well below the state averages every year. No sites in Depoe Bay were in the top 5 percent of historical accident sites listed in the SPIS records before June 27, 2008, and only two accidents were reported for the Collins Street and Highway 101 intersection in the five years before the Refinement Plan study.” Later, ODOT added small portions of Kargel’s deposition testimony to the summary judgment record, including Kargel’s statement that the noninclusion of the Collins Street intersection at the top of the SPIS rankings amounted to a decision that ODOT highway safety construction money “couldn’t have been spent for improvements” there. The portions of Kargel’s testimony that ODOT added to the record did not expand on the explanation of the operation of the SPIS-to-STIP process on which ODOT relied. Colip opposed the motion for summary judgment, arguing that ODOT’s theory was legally unsound and that, in light of the material she had submitted as part of the summary judgment record, ODOT had failed to establish, as a matter of law, that it was immune from liability for the alleged negligent omissions under ORS 30.265(6)(c). Among other things, Colip submitted: 1. An email from Kargel to another ODOT employee, Cantrell, describing the SPIS-to-STIP process, which included a statement that “[ojutside of the SPIS, other methods of flagging potential problem areas include citizen phone calls, local jurisdictional concerns, or planned construction projects where *656 safety reviews are done as part of the design process. Roadway improvements at these areas are prioritized along with SPIS sites based on Benefit/Cost calculations and available funding sources.” 2. ODOT’s Highway Safety Program Guide (September 2010), which includes, among other things, a statement that projects may become eligible for STIP funding by meeting one of three criteria—a top 10 percent SPIS ranking, 3 a cost-benefit ratio of 1.0 or greater, or justification by a so-called “risk narrative.” The guide also contains descriptions of the SPIS and of the method for determining a project’s cost-benefit ratio, along with a brief statement about risk narratives—that they “shall not be used to justify highway safety projects except for projects where crash trends may not be evident, such as bicycle or pedestrian improvements.” 3. An ODOT bulletin announcing that, effective September 20, 2007, ODOT had adopted the Highway Safety Engineering Quick Fix Program that provided ODOT regions with limited funds to implement low-cost (under $50,000) safety improvements without going through the STIP process. 4. A compilation of photographs showing that, within a few months of the accident that caused plaintiffs injuries, construction had been undertaken to create “bump outs” at the Collins Street intersection, presumably to improve the line of sight and presumably by ODOT. 5. Portions of Kargel’s deposition testimony in which she acknowledged that there were mechanisms outside of *657 the SPIS, such as requests from citizens, ODOT maintenance personnel, and other public bodies, that could trigger an investigation into whether a safety improvement is needed; that ODOT could make “minor” or “low-cost” improvements to portions of a roadway that are not listed in the STIP; that ODOT sometimes used money from maintenance and general budgets to fund such improvements; and that removing parking spaces or reconfiguring parking on Highway 101 at Collins Street would qualify as low-cost. Colip also submitted documentary evidence, including emails, meeting minutes, and planning documents suggesting that ODOT employees had been aware of the line-of-sight problems at the intersection for a number of years, had offered recommendations relating to the problem, and had participated in planning efforts, along with representatives of the city and county, that resulted in proposals for addressing the danger. Plaintiff had submitted many of the same materials in opposition to ODOT’s motion for summary judgment against his claims. As noted, the Court of Appeals reversed the trial court’s grant of summary judgment. It likened the circumstances to those in Vokoun v. City of Lake Oswego, 335 Or 19, 56 P3d 396 (2002). In that case, this court rejected the defendant city’s contention that its adoption of a capital improvement plan that did not include fixing the storm drain problem at issue in the case amounted to a policy decision for which it was immune under ORS 30.265(6)(c). Noting that the decision in Vokoun appeared to turn on (1) evidence that the city had a supplemental budget system for paying for repairs outside of the capital improvement plan and (2) the fact that there was no evidence that the city had considered using a supplemental budget to pay for the repairs, the Court of Appeals in the present case looked for similar alternatives to the SPIS-to-STIP funding process on which ODOT relied. The court concluded that the summary judgment record failed to establish “that all of the Highway 101 modifications in question were considered and rejected in the STIP process or that other available processes were used to decide to not make those changes. Turner, 270 Or App at 367. Summary judgment was not proper, the court concluded, because ODOT had failed to show that it had made *658 policy choices for which ODOT was entitled to immunity. Id. at 368. C. Arguments on Review Before this court, ODOT contends that the Court of Appeals’ decision erroneously makes the availability of discretionary-function immunity depend on the public body showing that it gave particularized consideration to every hazard and potential improvement at the site in question. ODOT asserts that, when a public agency has very broad responsibilities and limited resources, it should be entitled to discretionary-function immunity when it uses a non-particularized and global process, such as the SPIS-to-STIP process, to decide that it will not allocate its resources to lower-priority sites. That is so, ODOT argues, because decisions involving priority-setting and resource allocation are the essence of policy discretion and because denying a public body immunity for such decisions would result in utter paralysis: ODOT, for example, would have to spend all of its resources individually cataloging potential roadway hazards, leaving nothing to actually fix them. Plaintiff and Colip have different answers to ODOT’s argument. Plaintiff suggests that, in fact, discretionary-function immunity is not available for the kind of decision-making process that ODOT describes. Plaintiff observes that the crux of ODOT’s theory of budget-driven discretion is that, by employing a crash-history driven process to select hazardous sites for improvement, ODOT made a discretionary policy choice not to take action at sites that do not exhibit the required crash history. Yet, plaintiff notes, under this court’s cases, the state has a nondiscretionary duty to make state-owned highways reasonably safe for members of the public who use them in a manner that is consistent with their purpose. See generally Little v. Wimmer, 303 Or 580, 589, 739 P2d 564 (1987) (finding legislative intent to impose such a duty in ORS 366.205(2) and ORS 366.290(1)); Hughes v. Wilson, 345 Or 491, 497-98, 199 P3d 305 (2008) (holding that counties have a duty to make roads that they own and control reasonably safe for the general public). In light of that nondiscretionary duty, plaintiff argues, ODOT may have had discretion to choose the means by which to *659 make the roads reasonably safe, but it had no discretion as to whether to fulfill that duty. Plaintiff relies on Garrison v. Deschutes County, 334 Or 264, 274, 48 P3d 807 (2002) (“The decision whether to protect the public by taking preventive measures, or by warning of a danger, if legally required, is not discretionary; however, government’s choice of means for fulfilling that requirement may be discretionary.” (Emphasis in original.)). See also Hughes, 345 Or at 496 (“range of permissible choices does not * * * include the choice of not exercising care”). And because ODOT lacks discretion not to take action with respect to a particular hazard or site, plaintiff concludes, ODOT cannot claim discretionary-function immunity for such a decision. However, we do not reach plaintiffs argument, because Colip’s argument, which takes a different tack, is well taken. Colip calls into question whether ODOT has offered a factual showing that even makes relevant ODOT’s basic premise—that a public body’s budget-driven choice not to act on a particular hazard or site, because it has limited resources and decides that other hazards or sites are more important, is entitled to discretionary-function immunity. Colip’s argument to this court is that ODOT has failed to establish that, as a matter of law, the “global” SPIS-to-STIP process on which ODOT’s entire theory rests in fact was a policy decision not to fund or implement improvements at the Collins Street intersection. In that regard, Colip observes that a necessary element of ODOT’s theory is the idea that sites that do not rank at the top of the SPIS index do not get funded by ODOT. 4 Colip contends that the summary judgment record does not *660 support that element of the theory at all, much less support it as a matter of law. Colip points to evidence in the record—specifically, Kargel’s deposition testimony and her email to Cantrell—that shows that a site can be identified for consideration in the STIP process by means other than a top-five-percent ranking on the SPIS. Colip also notes that ODOT’s Highway Safety Program Guide, which is part of the record, shows that a project can actually qualify for STIP listing based on criteria other than a top-five-percent SPIS ranking—specifically, by having a “cost-benefit ratio of 1.0 or greater” or through a “risk narrative.” Finally, Colip argues that evidence in the record suggests that safety improvement funding for some “minor” or lower cost projects may be obtained outside of the STIP process. Colip particularly adverts to ODOT’s bulletin announcing the “Quick Fix” program and Kargel’s deposition testimony that “small amounts of money” may be available from maintenance and general budgets. Colip contends that at least some of the safety improvements mentioned as possibilities in the complaint and cross-complaint—specifically, parking changes and signage on Highway 101—might have qualified for such out-of-STIP funding. Colip also points to photographs in the record showing that improvements had been undertaken at the Collins Street intersection within a few months of the collision of plaintiffs motorcycle and Colip’s car, presumably by ODOT and presumably with funds that had not been obtained through the recent SPIS-to-STIP process. In light of all of that evidence, Colip argues, the factual underpinning of ODOT’s immunity claim is questionable, at best, and summary judgment should not have been granted. ODOT responds that Colip’s evidence of alternatives to the crash-history-based STIP process does not support the point that Colip wishes to make. According to ODOT, that is because the identified alternatives to the SPIS-to-STIP budgeting process all ultimately depend on the STIP process and the SPIS crash-history index that is its heart. In support of that contention, ODOT generally adverts to the 2010 Highway Safety Program Guide and the Quick Fix Program bulletin that Colip had introduced into the summary judgment record. It acknowledges that, for that purpose, the record is less than ideal. *661 In fact, as explained below, the record is far from ideal for purposes of ODOT’s summary judgment motion. For somewhat different reasons than those articulated by the Court of Appeals, we conclude that ODOT failed to establish a lack of genuine issues of fact for trial and that it was entitled to judgment based on discretionary-function immunity as a matter of law. D. Whether ODOT Was Entitled to Summary Judgment Part of ODOT’s claim of immunity is that, as a matter of discretionary policy, sites are included in appropriation requests and are funded through STIP based on sufficiently high SPIS rankings. If that were ODOT’s singular position, then logically, there would be no need for ODOT to respond to Colip’s evidence that there are ways, other than a high SPIS ranking, to identify sites for funding consideration in the STIP process. But ODOT makes a bolder claim: the SPIS-to-STIP process constitutes a global policy decision not to fund safety improvements at sites that are not highly ranked on the SPIS index. To succeed in making that claim, ODOT was obligated to respond to the evidence in the record that indicates that, for the kind of improvements suggested in plaintiffs complaint and Colip’s cross-complaint, (1) a high SPIS ranking was not the only path to inclusion in the biennial STIP listing and (2) inclusion in the STIP listing was not the only path to funding by ODOT. ODOT failed to do that. We turn, first, to the evidence that is relevant to the latter proposition. As noted above, to counter ODOT’s suggestion that funding for highway safety projects depends on their inclusion in the STIP list, Colip submitted several pieces of evidence (other than evidence that ODOT was aware of the hazard at the Collins Street exit): (1) an ODOT bulletin announcing a “Quick Fix” program that would provide some limited funding to address immediate highway safety concerns, without the necessity of going through the two- to six-year-long STIP process, and (2) Kargel’s deposition testimony acknowledging that ODOT can fund “minor” or “low-cost” highway safety improvements outside of the STIP, through maintenance or general budgets, and *662 that removing parking spaces or reconfiguring parking on Highway 101 at Collins Street would qualify as low-cost. ODOT does not attempt to meet the latter evidence with any evidence of its own, but it argues that ODOT’s immunity for a programmatic exercise of its policy discretion should not be lost merely because it makes some small amount of funding available for emergencies. For reasons that will become clear, we need not address that concern. As to Colip’s evidence of a “Quick Fix” alternative to STIP funding, ODOT points to the Quick Fix bulletin itself as showing that eligibility for that program also depends on SPIS crash-history rankings. We assume that ODOT refers to a statement in the bulletin that, to be chosen for funding under the Quick Fix program, projects shall “meet the guidance outlined in the ODOT Highway Safety Program Guide (B [enefit]/C [ost], SPIS top 5%, SIP 4 or 5 5 or Risk Narrative).” 6 If so, then ODOT is placing all of its chips on a single factual proposition—that, at least as they would apply to the improvements that Colip alleged should have been made, all of the alternative criteria for inclusion in the STIP list ultimately depend on SPIS crash-history rankings. If the evidence in the summary judgment record establishes the truth of that proposition as a matter of law, then Colip’s evidence that those alternatives exist does not undermine ODOT’s theory that its use of the SPIS-to-STIP process to select projects for funding was a policy decision that dictated nonaction with respect to lower-SPIS-ranked sites like the Collins Street intersection. As it turns out, the evidence in the record is insufficient for that purpose. The 2010 Highway Safety Program Guide appears to be the only evidence that addresses the STIP selection process in any detail. The Program Guide states: “Projects shall meet one of the following criteria for eligibility for Highway Safety Program funds: “* Positive Benefit/Cost (B/C) Ratio of 1.0 or greater *663 “* Top 10% Safety Priority Index System; or “* Justified by Risk Narrative [.] ” Only one of the three alternative criteria expressly relies on a high crash-history ranking. The question remains whether there is evidence that the other two criteria also rely in some fashion on a high SPIS ranking, or otherwise are made irrelevant to the Collins Street intersection and the improvements that plaintiff and Colip mentioned in their claims. The requirement of a positive “Benefit/Cost Ratio” of 1.0 or greater is explained briefly in the Program Guide. According to the guide, the Benefit/Cost Ratio is “the ratio of the economic value of the long-term reductions of target crashes to the estimated cost of the improvement.” The document describes how to determine costs and explains that the expected reduction in crashes (the “CRF”) should be drawn from a document maintained by ODOT that lists CRF for various types of improvements. The only mention of crash history data in the explanation is an admonition that “3-5 years of the most recent crash data available should be used for the analysis.” Although, based on that minimal explanation, we can imagine that it might be easier for a site with a high accident history to generate a higher Benefit/ Cost Ratio under those instructions, there is nothing in the explanation that suggests that a top-five-percent SPIS ranking or identification as a “high accident site” would be necessary to achieving a positive Benefit/Cost Ratio of at least one—particularly for some of the low cost improvements that Colip and plaintiff had mentioned in their claims. In short, the evidence in the record does not support ODOT’s suggestion that all possible routes to listing in the STIP ultimately were controlled by a policy decision to limit STIP funding to sites with top-five-percent or, at least, “high” crash-history rankings. The relevance of a site’s crash-history ranking to the “risk narrative” alternative is more obscure. The Program Guide—the only evidence in the summary judgment record that speaks to the question—describes the risk narrative alternative in the following terms: *664 “The Risk Narrative Form is a way to justify a project based on the safety hazard at a location that does not have available motor vehicle crash records or would typically not show evidence of a safety problem through crash records. Pedestrian and bicycle safety improvements are often justified by a risk narrative because they do not necessarily have significant crash history but have the potential for severe or fatal injury crashes. Safety projects for improving safety of motor vehicles should normally use the benefit cost analysis because they typically would have crash records associated with a location or segment of roadway. * * * Risk Narratives may not be used to justify roadway safety projects that would typically display crash trends but few or no crashes exist. A Risk Narrative (RN) should only be used when potential exists for high severity crashes and the nature of the crashes are such that they happen so sporadically that a crash history may not exist.” ODOT suggested at oral argument that risk narratives pertain to improvements for bicyclists and pedestrians. ODOT’s point would appear to be that, because Colip’s and plaintiffs claims address safety hazards at a location for which crash records are available and of a kind that would be made evident by crash records, the risk narrative alternative is irrelevant to the present question and cannot be used to place an improvement project for the crash site into the STIP. We disagree. It is at least arguable that the safety hazard at the Collins Street intersection was one for which there was a potential for high-severity crashes of a sort that “happen so sporadically that a crash history may not exist.” Whether to accept, as appropriate, a risk narrative pitched in that fashion would be a matter to be decided by an ODOT employee with the task of applying the policy set out in the Program Guide. And although it is possible, and perhaps even likely, that that employee would reject such a pitch, the decision to do so would not be a direct application of a high-level policy choice regarding the necessity of crash-history data but, rather, a determination that the conditions for application of that policy choice were not present. In the end, the material in the Program Guide describing the risk narrative alternative does not exclude that alternative from the range of possibilities for including the improvements at issue in the STIP list. *665 On this summary judgment record, then, there at least is a possibility that inclusion in the STIP could have been obtained through the risk narrative and Benefit/Cost Ratio alternatives to the SPIS crash-history ranking criterion and that those alternative criteria did not depend on the SPIS. ODOT therefore has not established, as a matter of law, that the kinds of improvements that Colip and plaintiff mentioned in their claims were automatically excluded from funding because of the Collins Street intersection’s failure to rank in the top five percent of the SPIS index or to be otherwise considered a “high accident site.” Insofar as that factual proposition is the lynchpin of ODOT’s theory of discretionary-function immunity, the theory fails. Because we have determined that ODOT has failed to establish under the requisite summary judgment standard a factual proposition that is essential to its affirmative defense of discretionary immunity based on the STIP selection process, we need not consider its contention that the existence of alternative funding sources for minor and emergency improvements (the Quick Fix program and maintenance and general budgets) should not undermine its discretionary immunity. Neither need we consider its attempts to distinguish the circumstances of the present case from those at issue in Vokoun, which, as noted above, the Court of Appeals relied on in reversing the trial court’s grant of summary judgment: ODOT’s argument with respect to Vokoun also relies on the factual proposition, which it has failed to prove as a matter of law, that ODOT’s global SPIS-to-STIP process constituted consideration of and a decision against any improvements at the Collins Street intersection. It may be that, on a more developed record, ODOT could have shown that the policy decision on which it relies in fact did exclude any projects at the Collins Street intersection from funding through ODOT. Or it may be that it could have established, in some other way, that global policy decisions based on its funding priorities necessarily had that effect. But the bottom line is that, on this record, a genuine issue of material fact exists as to whether sites that are not ranked in the top five percent of the SPIS index or are not considered “high accident sites” are excluded from ODOT highway safety funding, either through the STIP or *666 other highway safety funds. ODOT’s claim of discretionary-function immunity depended on such a finding, and the trial court erred in granting summary judgment to it based on that claim of immunity. The judgment of the circuit court is affirmed in part and reversed in part, and the case is remanded to the circuit court. The decision of the Court of Appeals is affirmed. Footnotes [1] The Court of Appeals also was unpersuaded by the city’s contention that its adoption of the aforementioned transportation plan and refinement plan amounted to policy choices that immunized it from liability for its inaction with respect to the hazard at the Collins Street intersection. The court concluded that the transportation and refinement plans constituted evidence that the city had made a decision to do something to mitigate the hazard at the intersection, but, in the absence of evidence that it had taken action to effectuate that decision, discretionary immunity was not available. 270 Or App at 368-70. The Court of Appeals thereby rejected the trial court’s reasoning that evidence of the participation of all three government bodies in an ongoing planning process to mitigate the hazard was sufficient to establish discretionary immunity as a matter of law for the actions and omissions alleged in the complaint and cross-complaint. Id. [2] As noted above, ODOT raised discretionary-function immunity in both of its motions for summary judgment'—one against plaintiff and one against defendant/cross-plaintiff Colip. Although the focus of our analysis rests on the latter motion, which the trial court actually decided on the basis of ODOT’s claim of discretionary-function immunity, given the trial court’s reasoning and the Court of Appeals decision, our review has implications for ODOT’s summary judgment motion against plaintiff as well, and plaintiff therefore has participated in the arguments. [3] Kargel’s affidavit and deposition testimony refer to the top five percent of the SPIS as a criterion for funding eligibility, while the 2010 ODOT Highway Safety Program Guide that Colip submitted into the record refers to the criterion as the top 10 percent of the SPIS. The difference seemingly arises from the different time period that the Program Guide speaks to (notably, the 2007 Quick Fix Program bulletin that Colip submitted into the record refers to the criteria in the Highway Safety Program Guide as “SPIS top 5%”). In the trial court, ODOT did not challenge the relevance of the 2010 Highway Safety Program Guide to its theory of discretionary-function immunity as it pertains to its actions in 2008 (the time of the accident) and earlier. In fact, it chose to supply the 2010 Program Guide, rather than an earlier version, in response to plaintiffs discovery request for “documents describing the factors, procedures or analysis *.. by which defendant selects highway safety construction projects.” [4] In so arguing, Colip sets aside ODOT’s additional mention of “other high accident rated sites based on a eost/benefit analysis.” We agree with that approach: ODOT never explains how that alternative fits into its SPIS-to-STIP prioritization theory, and, as discussed below, the scant evidence in the summary judgment record that speaks to the cost-benefit option does not support the idea that the option depends on SPIS crash-history ranking. An additional point that is worth making at this juncture is that, when Kargel mentioned “other high accident rated sites based on a cost/benefit analysis,” she was not referring to some entirely separate “cost-benefit” metric. Her declarations that were specific to the Collins Street intersection only pertained to the area’s crash history, as recorded in the SPIS. [5] The SIP index mentioned here apparently was relevant, at one time, in ODOT’s process for prioritizing highway safety construction projects. [6] We have found no other material in the Quick Fix Bulletin or in the summary judgment record as a whole that appears to be relevant to ODOT’s claim.