OPINION
LILLEHAUG, Justice.RDNT, LLC asks us to hold that the City of Bloomington’s decision to deny RDNT’s conditional use permit application was unreasonable, arbitrary, and capricious, and to hold that the City did not properly consider RDNT’s proposed traffic-mitigating conditions. We hold that the City’s decision was not unreasonable, arbitrary, or capricious, as the City based its decision on a legally and factually sufficient ground: that the proposed use would be injurious to the surrounding neighborhood or otherwise harm the public health, safety, and welfare. We also hold that the City’s determination that RDNT’s proposed efforts to mitigate traffic were insufficient was not unreasonable, arbitrary, or capricious.
I.
RDNT, LLC (“RDNT”) owns the Martin Luther Care Campus (“Campus”), lo*73cated in Bloomington. The Campus consists of two buildings: the Martin Luther Care Center and Meadow Woods Assisted Living. The Campus provides a variety of services, including assisted living, memory care, skilled nursing, adult day care, and transitional care.
On September 27, 2011, RDNT submitted an application to the City for a conditional use permit. In its application, RDNT sought to expand its existing assisted living services by adding a third building to the Campus. RDNT stated that the expansion would allow those served by its existing transitional care unit to transfer into the assisted living units, thereby allowing them to “age in place.”
At the time, the Campus consisted of 137 units in its skilled nursing facility and 117 units in its assisted living facility. The proposed addition would be three stories tall and contain 67 “catered living units,” increasing the total units from 254 to 321: a 26 percent increase. It would also increase the staff from 186 to 202 employees: an 8 percent increase. And it would increase the total building square footage from 198,209 square feet to 321,264 square feet: a 62 percent increase.
On November 3, 2011, RDNT presented its application to the Bloomington Planning Commission in a meeting open to public comment. Numerous citizens spoke about the proposed expansion, with many voicing concerns about increased traffic. The Planning Commission unanimously voted to recommend denial of the conditional use permit application. The Planning Commission adopted the view of its staff report that the proposed expansion would violate the City’s comprehensive plan because: 1) it is not adjacent to an arterial or collector street; 2) it is not in close proximity to transit, amenities, and services; and 3) it would not preserve the character of the surrounding low density, single family neighborhood. The Planning Commission also adopted the staffs view that the proposed use would violate the City’s conditional use permit ordinance because it would be injurious to the surrounding neighborhood or otherwise harm the public health, safety, and welfare. The staff based its views on estimated increases in traffic and on the size, density, and design of the proposed building.
The City Council met on November 21, 2011, to consider the application. Among other materials, the City Council reviewed traffic studies from two different experts estimating the future traffic volume that would be generated by the proposed expansion.
SRF Consulting Group, Inc. (“SRF”), hired by the City, conducted the first study. Using data collected on its own and by City staff, SRF calculated that the Campus generated 1,145 trips on an average day, resulting in 4.50 trips per bed. That figure exceeded the standard trip generation rate for similar facilities, which is 2.66 trips per bed. SRF relied on the 4.50 trips per bed rate in order to provide a “conservative estimate.” SRF then examined the extent to which the proposed expansion would increase traffic. Using the “actual (collected) trip generation rates,” SRF determined that the expansion would increase the daily number of trips from 1,145 to 1,447: a 26 percent increase. SRF concluded that the Campus generated trips on the higher end compared to similar facilities. SRF also concluded that “existing neighborhood roadways and intersections have sufficient capacity to accommodate the additional vehicles expected from the ... expansion and no additional roadway improvements would be necessary to accommodate the expansion as proposed.” SRF noted that delays at East Old Shakopee Road and *74East 98th Street would likely increase, but the increase would be “minimal.”
URS Corporation (“URS”), retained by RDNT, conducted the second study. In evaluating the effect of the expansion on the number of trips generated, URS acknowledged that SRF’s collected trip generation rate was calculated using an “accepted method.” However, URS contended that SRF based its rate on incorrect assumptions. URS contended that the new addition would share trips with the current site and would thus not increase traffic at the same rate that the existing use generated. URS further asserted that because the new facility would offer a “lower intensity of care,” similar to the existing assisted living facility, it would require fewer employees and trips than a more intensive-care facility, like the existing skilled nursing facility. ' Instead, URS determined that the industry standard rate of 2.74 trips per bed should apply to the new expansion. This would result in an additional 184 daily trips, for a total of 1,329 daily trips: a 16 percent increase. URS also evaluated the Campus’s existing Transportation Demand Management Program (“TDMP”), which included a variety of measures designed to reduce the number of new and existing trips generated by the Campus. URS concluded that a more robust TDMP could further reduce the number of trips generated by the Campus by up to 70 trips per day. URS also determined that working with major vendors could reduce truck trips by six to eight trips per day.
In addition to the expert studies, the City Council received information from the neighborhood. Those in favor of the expansion highlighted the individual and community benefits of assisted living for the elderly. Those in opposition focused primarily on the effect that the Campus’s existing traffic had on noise, safety, home values, and the general character of the neighborhood. Public comments given at the City Council meeting echoed the written concerns, with the primary emphasis on traffic issues.
By a vote of four to three, the City Council passed a resolution to deny RDNT’s application for a conditional use permit. The resolution set forth four.reasons for the denial. The first three reasons related to conflicts with different comprehensive plan provisions. The fourth reason related to the City conditional use permit ordinance, which requires that the “proposed use will not be injurious to the surrounding neighborhood or otherwise harm the public health, safety and welfare.” Bloomington, Minn., Code of Ordinances ch. 21, art. V, div. A, § 21.501.04(e)(5) (2014).1 The City Council found that the increase in square footage rendered the expanded Campus “incompatible with the scale and character of the surrounding low density, single family neighborhood.” The City Council found that most structures in the neighborhood would be one-fifth or less than the size of the new addition. It also cited increased traffic volumes, projected to total between 1,377 and 1,447 daily trips, as injurious or otherwise harmful. The City Council further found that the traffic volume would be “over three and one half times the average daily trips” than if the 13-acre site were instead developed to a low density of three units per acre. And it found the various traffic concerns submitted by the public to be “credible and consistent with the traffic studies presented and staff analysis of the application.” Finally, the City Council found RDNT’s TDMP to be insufficient to *75avoid the injury, given the location and nature of the Campus.
RDNT filed a complaint and petition for alternative writ of mandamus with the Hennepin County District Court. On cross-motions for summary judgment, the district court granted summary judgment to RDNT and reversed the denial of RDNT’s application. The district court held that, for each of the four reasons given, the City “misapplied certain standards, misrepresented the impact of certain studies, and appeared to ignore evidence to the contrary.” Specifically, as to the fourth reason, the district court held that the record was insufficient to support a finding that the proposed use would injure the neighborhood or harm the community. The district court criticized the City’s reliance on the SRF study and the generalized neighborhood opposition.
The City appealed. In an unpublished opinion, the court of appeals reversed, holding that the City appropriately exercised its discretion. RDNT, LLC v. City of Bloomington, No. A13-0310, 2014 WL 30382 (Minn.App. Jan. 6, 2014). Of the four reasons given by the City for its decision, the court of appeals held that three were factually and legally sufficient.2 Id. at *2-9. On the fourth reason identified by the City, involving injury to the neighborhood and harm to the public health, safety, and welfare, the court of appeals held that the City ordinance from which that standard was derived was legally sufficient, and that the City was within its discretion in finding that the proposed use violated this ordinance, based on the square-footage and traffic evidence. Id. at *9.
We granted RDNT’s petition for review.
II.
RDNT argues that each of the City’s four reasons for denying RDNT’s conditional use permit application was arbitrary and capricious. We limit our consideration to the City’s fourth reason: that the proposed use would violate subdivision (e)(5) of the City’s conditional use permit ordinance.3
“[T]he interpretation of an existing ordinance is a question of law for the court.” Frank’s Nursery Sales, Inc. v. City of Roseville, 295 N.W.2d 604, 608 (Minn.1980). We review a question of law de novo. Alpha Real Estate Co. of Rochester v. Delta Dental Plan of Minn., 664 N.W.2d 303, 311 (Minn.2003).
We will reverse a governing body’s decision regarding a conditional use permit application if the governing body acted unreasonably, arbitrarily, or capriciously. Schwardt v. Cnty. of Watonwan, 656 N.W.2d 383, 386 (Minn.2003).4 There are two steps in determining whether a city’s denial was unreasonable, arbitrary, or capricious. First, we must determine if the reasons given by the city were legally *76sufficient. C.R. Invs., Inc. v. Vill. of Shoreview, 304 N.W.2d 320, 325 (Minn.1981). Second, if the reasons given are legally sufficient, we must determine if the reasons had a factual basis in the record. Id.
A.
We first address whether the City’s conditional use permit ordinance is legally sufficient. The City’s ordinance states in relevant part: “The following findings must be made prior to the approval of a conditional use permit: ... (5) The proposed use will not be injurious to the surrounding neighborhood of otherwise harm the public health, safety and welfare.” Bloomington, Minn., Code of Ordinances ch. 21, art. V, div. A, § 21.501.04(e)(5) (2014). We have long held that a city may deny a conditional use permit application if the proposed use endangers “the public health or safety or the general welfare of the area affected or the community as a whole.” Zylka v. City of Crystal, 283 Minn. 192, 196, 167 N.W.2d 45, 49 (1969). However, we have also held that the “absence of more express standards makes denial of a special-use permit more, not less, vulnerable to a finding of arbitrariness.” Hay v. Twp. of Grow, 296 Minn. 1, 6, 206 N.W.2d 19, 22-23 (1973).
Thus, we hold that the City ordinance is legally sufficient, but will examine the factual basis for the City’s findings more closely than we would under a less subjective standard.
B.
Second, we address whether the City had a reasonable factual basis to determine that the proposed use would injure the surrounding neighborhood or otherwise harm the public health, safety, and welfare. We hold that it did.
The City and RDNT offer conflicting evidence about the effect the proposed use would have on traffic. Upon review of a city’s decision, our function is “not to weigh the evidence, but to review the record to determine whether there was legal evidence to support the zoning authority’s decision.” Barton Contracting Co. v. City of Afton, 268 N.W.2d 712, 718 (Minn.1978). With expert witnesses, we do not “attempt to weigh the credibility of conflicting experts, but instead review the record to ensure that the decision ... had support in the record.” Billy Graham Evangelistic Ass’n v. City of Minneapolis, 667 N.W.2d 117, 124 (Minn.2003). In other words, courts should ordinarily defer to a city’s judgment on conflicting evidence. See White Bear Docking & Storage, Inc. v. City of White Bear Lake, 324 N.W.2d 174, 176 (Minn.1982).
Even so, it is not necessary for us in this case to defer to the City’s projections to determine whether it had a reasonable factual basis. The City estimated that the daily number of trips would increase after the expansion from 1,145 to somewhere between 1,377 and 1,447, while RDNT estimated that the number of trips would be 1,259, taking the TDMP into account. Even by RDNT’s projection, the expansion would add over 100 daily trips. Further, the City’s engineer estimated that average traffic counts for a residential street were between 300 to 500 trips, and that the public tends to complain once traffic increases to 1,000 trips per day on such a street. Thus, regardless of which estimate is more accurate, there is a factual basis in the record for the City to find that the proposed expansion would increase traffic on already busy streets.
The evidence relied on by the City to find that the increase would injure the neighborhood is distinguishable from the evidence we held to be insufficient in C.R. *77Investments v. Village of Shoreview, 304 N.W.2d 320, 325 (Minn.1981). In that case the only evidence of a traffic control problem on a certain road was “the statement of one council member that he had been told of a problem existing at one intersection and his opinion that additional housing units might aggravate that problem.” Id. Here, neighbors gave concrete testimony about how the increase would exacerbate existing traffic conditions. For instance, one neighbor wrote about vehicles driving through crosswalks near the school, even though the crossing guards had their flags out. Another neighbor wrote about observing vehicles that sped and made U-turns. Yet another neighbor wrote about the incredible amount of “traffic and noise” due to the large number of delivery trucks, emergency vehicles, shuttle buses, passenger cars, and garbage vehicles. Thus, RDNT’s argument that the City relied on vague concerns for public health and welfare is simply unfounded: the City had in hand multiple traffic studies, the City engineer’s testimony regarding specific data, and detailed factual complaints from the neighborhood.
Still, RDNT argues that because the streets are not near capacity, as both the City’s engineer and SRF acknowledge, the City had no factual basis to deny the application. For this argument, RDNT relies on Chanhassen Estates, in which we held that the evidence was insufficient to reverse the grant of a conditional use permit where the neighborhood organization only offered “non-specific testimony that the proposed [use] poses potential traffic hazards at the intersection,” while the city offered “the city engineer’s testimony that the intersection could handle the anticipated traffic.” Chanhassen Estates Residents Ass’n v. City of Chanhassen, 342 N.W.2d 335, 340 (Minn.1984). Chanhas-sen Estates is distinguishable. In that case, the city engineer concluded that the intersection could handle the traffic, and we deferred to that conclusion in declining to reverse the city’s decision to grant a conditional use permit. Id. Here, although the City’s engineer concluded the streets were not at capacity, the City relied on specific evidence — traffic studies, average street numbers, and neighborhood testimony — to conclude that the proposed use would nonetheless injure or otherwise harm the neighborhood. Not unreasonably, the City determined’ that street capacity alone was not dispositive as to whether an increase in traffic injures the neighborhood or otherwise harms the public health, safety, and welfare. The fact that a street could physically handle more traffic does not determine whether the neighborhood or the public could handle more traffic. To paraphrase one of the City’s planners: this is not a capacity issue, it is a livability issue. On that issue, we cannot say the City acted unreasonably, arbitrarily, or capriciously.
Because of the traffic studies, the City engineer’s testimony, and the neighborhood testimony, we hold that the City had a sufficient factual basis to determine that the increase in traffic would injure the neighborhood or otherwise harm the public health, safety, and welfare.5
*78III.
RDNT also contends that the City unreasonably, arbitrarily, and capriciously determined that RDNT’s proposed mitigation efforts would be insufficient to alleviate the traffic issue. We disagree.
Minnesota Statutes § 462.3595 provides: “Conditional uses may be approved by the governing body ... by a showing by the applicant that the standards and criteria stated in the ordinance will be satisfied.” Minn.Stat. § 462.3595, subd. 1 (2014) (emphasis added). Thus, the burden was on RDNT. to show that it could satisfy the standards specified by ordinance. If a conditional use permit applicant demonstrates to the governing body that imposing a reasonable condition would eliminate any conflict with the ordinance’s standards and criteria, it follows that the governing body’s subsequent denial would be arbitrary. See Zylka v. City of Crystal, 283 Minn. 192, 196, 167 N.W.2d 45, 49 (1969) (“A denial would be arbitrary ... if it was established that all of the standards specified by the ordinance as a condition to granting the permit have been met.” (footnote omitted)).
RDNT argues that the City should have suggested or imposed mitigating conditions. We have held that a city’s decision to deny a conditional use permit was arbitrary in part because “there was no attempt made, either by the opponents or the council, to suggest or impose conditions” that would mitigate problems with the development. Minnetonka Congregation of Jehovah’s Witnesses, Inc. v. Svee, 303 Minn. 79, 85-86, 226 N.W.2d 306, 309 (1975). But we decided Svee before the Legislature enacted section 462.3595, Act of Mar. 22, 1982, ch. 507, § 24, 1982 Minn. Laws 587, 594 (codified at Minn.Stat. § 462.3595 (2014)), which places the burden on the applicant to satisfy the standards and criteria in the ordinance. To the extent there is any conflict between Svee and the later-enacted statute, the plain language of the statute controls.
Regardless of Svee, RDNT relies on C.R. Investments to argue that the City did not adequately consider RDNT’s proposed mitigating conditions. See C.R. Invs., Inc. v. Vill. of Shoreview, 304 N.W.2d 320 (Minn.1981). In that case, the village council had been informed that a traffic hazard could be eliminated by furnishing turn-around areas in the driveways. Id. at 325. The village council ignored that proposed reasonable condition. Id. Taking the ignored reasonable condition into consideration, we found no evidence warranting an inference that the traffic aggravation would be “substantial,” and thus held the village’s decision to be arbitrary. Id.
Unlike the village in C.R. Investments, which' ignored the applicant’s reasonable proposal,, see id., the record before us shows that the City adequately considered the proposed mitigating conditions in several ways. In rejecting RDNT’s application, the City concluded that the “minimization of the trip volume by the proposed Traffic Demand Management Plan is insufficient to avoid injury to' the public health, welfare and safety of the residential neighborhood.” In SRF’s expert report, heavily relied on by the City, SRF considered the TDMP. SRF admitted, taking the addition into account, that the TDMP had “the potential to reduce the facility trips by approximately 70 trips per day.” However, SRF cautioned that the “effectiveness of these strategies is difficult to quantify without post implementation data to support this forecast reduction; caution should be demonstrat*79ed when estimating the actual impact to trip reduction.” SRF also warned that “TDMPs typically tend to be the most effective in central business districts or in heavily transit-oriented developments versus the suburb-type development that this is.”6 Even if URS’s estimates regarding traffic generation and the TDMP’s effectiveness were accurate, the expansion would still add over 100 daily trips. Thus, the City had a reasonable factual basis to determine that the proposal would not alleviate the traffic concerns.
Accordingly, based on the record, the City did not act unreasonably, arbitrarily, or capriciously when it denied RDNT’s conditional use permit application.
Affirmed.
DIETZEN, J., took no part in the consideration or decision of this case.. Available at http://bloomingtonmn.gov/clerk/ city-charter-and-code-ordinances.
. The court did not address one of the comprehensive plan provisions considered by the City. RDNT, 2014 WL 30382 at *8.
. Because we hold this ground to be legally and factually sufficient, we do not need to address the three other grounds upon which the City based its decision. See Hubbard Broad., Inc. v. City of Afton, 323 N.W.2d 757, 765 n. 4 (Minn.1982) ("Not all of the reasons stated need be legally sufficient and supported by facts in the record.”).
.Since zoning laws are a restriction on the use of private property, a landowner whose application for a conditional use permit has been denied has a lighter burden than one who challenges approval of a permit. Bd. of Supervisors of Benton Twp. v. Carver Cnty. Bd. of Comm'rs, 302 Minn. 493, 499, 225 N.W.2d 815, 819 (1975).
. We appreciate the participation of, and the information furnished by, the amici who serve or represent some of the most vulnerable Minnesotans. While the amici explain cogently the difficulties providers encounter in siting, constructing, and operating care facilities, our task in this case, like any other, is not to make legislative policy but to interpret and apply existing statutes, ordinances, and precedents. See Dahlin v. Kroening, 796 N.W.2d 503, 508 (Minn.2011) (considering competing public policy arguments related to judgment renewals and holding that "policy-related issues are best left to the Legislature. When interpreting the statutes, it is our role to rely on what the Legislature intended over what *78may appear to be supported by public policy.”).
. One of the council members expressed similar concerns:
I appreciate all the effort that you're talking about trying to minimize, and I just — I’m looking at a bus stop six blocks away in the Minnesota winter going to a nursing home. I just don’t see that some of those measures are going to have a big enough impact to bring these numbers down to anywhere where they're acceptable.