OPINION
MOORE, Justice.This case centers around the Department of Community and Regional Affairs’ determination of the “population” of municipalities for state revenue-sharing and tax-limitation purposes. Specifically at issue is the department’s method of counting remote site workers on the North Slope. These workers present a problem because of their unique working situation. Typical work schedules may be one week on, one week off; two on, one off; or four on, two off. Because of the remoteness of the North Slope, employees often reside in other municipalities or even in other states during their time off. The central question in this case is where these workers should be counted.
This is an appeal from the superior court’s decision that the department’s 1983 determinations of population were a rational exercise of its discretion. We affirm.
I. STATUTORY BACKGROUND1
A) The Revenue-sharing Statute
Alaska Statute 29.88.0102 directs the Department of Community and Regional Affairs to compute an “equalization entitlement” for determining each municipality’s share of the amount of state revenues apportioned by-the legislature to the equalization account each year. See AS 29.88.035. The entitlement is based “on the population, relative ability to generate revenue, and local tax burden of the taxing unit.” *169It is to be calculated by the application of a formula which essentially multiplies the population of the municipality by the hypothetical local property tax rate if all local revenues were generated from property taxes. Alaska Statute 29.88.015 addresses the determination of population for insertion in this formula:
(a) For purposes of this chapter, the population of a taxing unit shall be determined annually by the latest figures of the United States Bureau of the Census or other population data which, in the judgment of the department, is reliable.
(b) The population of the taxing unit includes the population of any military reservation which is a part of the taxing unit.
“Population” is not- further defined anywhere in chapter 88.3
B) The Tax-limitation Statute
Alaska Statute 29.53.045 4 allows municipalities to levy taxes on specified oil and gas property, up to one of two alternative limits: (1) a municipality may tax the full value of this property at a rate such that its total tax revenues do not exceed “$1,500 a year for each person residing within its boundaries”; or (2) a municipality may tax the full value of a portion of this property so long as the value of that portion, when combined with the value of property otherwise taxable by the municipality, does not exceed the product of 225% of the state’s average per capita property value multiplied by “the number of residents of the taxing municipality.” For purposes of this case, the salient feature of both of these limits is that municipal population is a factor in their calculation. Alaska Statute 29.53.045(e) provides:
For purposes of this section, population shall be determined by the commissioner of community and regional affairs based on the latest statistics of the United States Bureau of the Census or on other reliable population data....
*170Alaska Statute 29.53.0505 applies the same two alternate measures of tax-limitation to a municipality’s general ability to tax, repeating verbatim the directive regarding the determination of population.
II. FACTS AND PROCEEDINGS
A) Population Counts Prior to 1980
The U.S. Census Bureau counts people at their “usual residence,” generally construed as the place where a person lives and sleeps most of the time. For the vast majority of the general population, the decision as to what constitutes one’s “usual residence” is left to the individual, based on guidelines in the census questionnaire. However, the U.S. Census Bureau has historically counted people who are quartered in workcamps at the place where they spent most nights during the week that the census was taken (“the four-night rule”).
Prior to 1980, the Alaska State Department of Community and Regional Affairs based its population estimates on the 1970 federal decennial census, plus “other reliable data” submitted by the municipalities (usually based on school enrollments and utility hookups). The department’s historical practice has been to use the same population figures for both revenue-sharing and tax-limitation formulas. Since 1976, it has made population determinations pursuant to its own regulations for revenue-sharing purposes, but it has never promulgated regulations regarding tax-limitation calculations.6
B) 1980 Population Counts
In response to requests from oil companies, municipalities, and the governor of Alaska, in 1980 the U.S. Census Bureau counted remote site oil workers in Alaska at their “home” addresses. Although the record is not entirely clear, it appears that “the four-night rule” was applied to workers in group living quarters in other states. The Bureau believed that this approach would better implement its policy of counting people at their usual residences. Its justification for changing its methods only for Alaska was that most work sites are not as remote as those in Alaska.
In 1980, the state legislature enacted the revenue-sharing plan currently at issue, in-*171eluding for the first time the concept of an equalization formula. In order to implement the new plan, the department issued emergency regulations indicating that it would base its population counts on the 1980 U.S. Census figures. Under these regulations, the, department initially approved a population estimate of 4,610 for the North Slope Borough. Acting on an appeal from the borough and advice from the state attorney general’s office, the department later adjusted this count to 9,234, returning to the previous practice of including remote site oil workers in the borough’s population count.
C) 1981 Population Counts
The department was disturbed to find that its 1980 population figures produced a statewide total that was substantially higher than the U.S. Census Bureau’s 1980 count for Alaska. In 1981, in order to insure that its figures would continue to be accepted by the federal government for the distribution of federal funds, the department issued a new set of permanent regulations regarding the determination of population for revenue-sharing purposes. These regulations, which are still in effect, state that the department will base its estimates on the latest figures of the U.S. Census Bureau, subject to adjustment upon its own motion or at the request of another state agency or municipality. A municipality that submits a request for adjustment must substantiate its figures with a census of its own or with estimates, either of which may count only those people who would be considered residents by the U.S. Bureau of the Census. 19 AAC 30.041.7 The regulations include the following definitions:
(3) “census” means a complete and accurate enumeration of the permanent residents and housing units in an area;
[[Image here]]
(9) “permanent resident” means a person whose primary place of residence is within the corporate limits of [the] municipality ...;
(10) “primary place of residence” means the place where a person usually sleeps on a weekly basis; if a person has more than one place of residence, “primary place of residence” means the place where the person sleeps more often during the calendar year;
19 AAC 30.150.
Pursuant to these regulations, the department counted people at the place they reside during the majority of the year (“the 51% rule”). This approach produced a 1981 population count of 7,098 for the North Slope Borough.
D) Initial 1982 Population Counts
In 1982, the legislature passed SB 180 which included provisions specifically di*172recting the department to include all persons working at isolated job sites in a municipality when determining population for both revenue-sharing and tax-limitation purposes. Governor Hammond vetoed this bill for a variety of reasons, one of which was advice from counsel that the bill’s technique for determining population might result in substantial losses in federal funding to the state.8 In his veto message, the governor expressly directed the department
to draft regulations on the provisions of state assistance to local governments in a manner which compensates more equitably those communities impacted by seasonal, temporary and isolated workers. Minimally, I would hope that in the short term we could at least “hold harmless” the North Slope Borough, which otherwise stands to lose about $2 million in revenues from the amount they received last year. All other municipalities would receive, I’m told, increases. Accordingly, I would hope that all other municipalities, which rose in violent protest over the prospects of revenue losses to themselves were SB 180 to become law, would be equally concerned about revenue losses incurred by other municipalities through this measure’s veto.
2 Senate Journal 1791 (2d Sess.1982).
In response to the governor’s directive, the department investigated the feasibility of a variety of approaches for estimating population. It considered counting “de fac-to” populations — that is, the total number of people physically present in each municipality on the date of the census — but rejected this approach because of the absence of benchmark data against which to test the accuracy of the initial de facto counts and the necessity of hiring additional personnel to implement such a program. The department also sent a questionnaire to each municipality requesting information on the impact of seasonal and temporary residents. Unfortunately, the responses were not very helpful; most municipalities could do no more than guess at the impact of various transient groups.
The department also commissioned a study of population impact by the Institute of Social and Economic Research (“ISER”). The ISER report concluded that the most accurate method of counting population for the purpose of assessing impact on municipal services would be to apportion individuals among all the different locations where they spend time, on the basis of the amount of time the individual spends in each location, multiplied by an index number which measures the relative amount of services the individual consumes there. However, the department rejected the approach recommended by ISER because of its complexity. It was concerned about the massive amount of data which would be required, the high data collection costs, and it feared that any weighting system would be vulnerable to criticism that it was arbitrary, subjective and politically motivated.
In order to meet impending statutory deadlines, the department finally adopted emergency regulations for measuring population for revenue-sharing purposes. Like the governor’s veto message, the department’s finding of emergency specifically expressed concern for the North Slope Borough:
[The] 1980 census procedure and its subsequent application to the revenue sharing program improperly caused the North Slope Borough to lose some revenues under the State Revenue Sharing and Municipal Assistance programs but more significantly unduly limited revenues under the related local tax limitation provisions found in AS 29.53.050(b). These constraints produced serious unanticipated revenue shortfalls for the North Slope Borough and has [sic] an indirect effect on its ability to sell bonds and meet future financial obligations. Therefore, to prevent a further reduction of the financial stability of the North Slope Borough to carry out essential municipal programs, these emergency regulations are adopted.
*173The emergency regulations essentially reinstated the method which was used by the U.S. Census Bureau prior to 1980. In order to compute the set of figures prepared in accordance with these regulations, the department started out with a base count for each municipality determined under “the 51% rule” embodied in the 1981 regulations. Then it applied “the four-night rule” to remote site oil workers only.9 To determine the North Slope Borough’s population, it took the average of a) the borough’s estimated 1982 population based on “the 51% rule” (7,552) and b) the actual number of persons present during a local 1981 census paid for by the borough and conducted by a private contractor under the Department of Labor’s supervision (11,-253). This averaging method produced a population count of 9,402. Estimates from this local 1981 census were also used to determine how many people to subtract from other municipalities’ “51%” base counts because of their reallocation to the North Slope Borough.10
The set of figures produced by this calculation was certified by the department as the final 1982 population counts.11
E) The Lawsuit (Round One)
Shortly after the department’s release of the 1982 determinations, four municipalities — Mat-Su, Anchorage, Kenai Peninsula and Fairbanks (collectively, “Mat-Su”) — requested a declaratory judgment on the meaning of the term “population” in AS 29.88.015 and asked the court to order the department to recertify new 1982 figures on the basis of this judicial interpretation.
The superior court held that the emergency regulations were invalid because no real emergency existed. However, the court also held that the department has broad discretion under the applicable statutes and the 1981 regulations to establish a definition of population. Accordingly, the superior court believed that judicial review should be limited to determining whether the department had a rational basis for its selection of computation methods and for judging its data to be reliable.
Applying the rational basis standard, the court struck down the 1982 population counts on two main grounds: (1) there was no rational basis for assuming that the averaging method used for computing the North Slope Borough’s population would fairly estimate the results of the four-night rule; and (2) while it might have been rational to apply “the four-night rule” to the North Slope Borough alone if oil slope workers were unique as a stable, year-round workforce, there was no such evidence before the court. The state demographer testified that the population of the *174oil camps varies considerably throughout the year. The court therefore held the special treatment accorded to the North Slope unfairly discriminates against other boroughs which have significant seasonal and temporary populations.
The court remanded the matter to the department “to recertify new 1982 population figures in its discretion in any lawful manner.”
F)Revised 1982 Population Counts
In response to the court’s order, the department ultimately certified a new set of 1982 figures based on the 51% rule12 for all municipalities except the North Slope Borough. For the borough, the department certified two figures: a revenue-sharing count of 5,118 based on the 1980 U.S. Census method and a tax-limitation count of 10,427 which included the average remote site workforce. The latter number was calculated by adding to the revenue-sharing figure the total number of mining and construction jobs in the remote site work-camps, based on employment statistics kept by the Department of Labor. If, instead, the department had applied the 51% rule to the North Slope Borough as it did to all the other municipalities, its population count would have been 7,552.
The decision to use two figures for the borough was a compromise in a sense. On the one hand, the department wanted to eliminate all double counting for revenue-sharing purposes on the theory that revenue-sharing is a strict per capita concept. On the other hand, the department wanted to avoid using unrealistically low population figures for the North Slope Borough that would hinder its ability to support itself through local taxation. The department wanted to take into account the impact of the remote site workforce on municipal services, and it wanted to prevent a sudden and disruptive reduction in the amount of funds available to the borough.
None of the parties to the litigation appealed the revised 1982 figures or objected formally to the use of two figures for the North Slope Borough in 1982.
G) 1983 Population Counts
For the 1983 counts, the department once again certified a single set of figures based on the 51% rule for all municipalities except the North Slope Borough. For the borough, it certified two population figures: a revenue-sharing count of 7,721 based on the 51% rule and a tax-limitation count of 10,171 which included the average remote site workforce.
In order to calculate the tax-limitation figure, the department started with an estimate based on the 1980 U.S. Census method. Then it obtained employment data on the mining and construction industries for the entire borough from the Department of Labor. It added to the initial figure 100% of the annual average number of mining employees and half of the annual average number of construction employees, on the theory that the other half had already been counted in the base population.
The mining and construction statistics used for the 1982 population counts had been based on the Prudhoe Bay/Deadhorse area alone rather than the borough as a whole. Apparently, Commissioner Mark Lewis, who did the calculations in 1982, did not inform his successor, Commissioner Jeff Smith, that more detailed breakdowns of the borough were available. The deputy director of the division of municipal and regional assistance, Douglas Griffin, estimates that if the department had used more precise employment data in 1983 as it did in 1982, the 1983 tax-limitation figure for the borough would have increased by more than 1000.
H) The Lawsuit (Round Two)
Following the department’s release of the revised 1982 figures and the 1983 figures, Mat-Su filed an amended complaint once again requesting a declaratory judgment establishing the statutory limits on *175the term “population” and declaring invalid any determination of population in violation of those limits.13
Ruling on cross-motions for summary judgment, the superior court issued a memorandum opinion holding that the 1983 population figures certified by the department were reasonably based on reliable data. It is Mat-Su’s appeal from that decision which is now before us.
III. DISCUSSION
A) The Department’s Population Determinations
1) Standard of Review
There are two standards under which this court has reviewed agency interpretations of statutory terms. The reasonable basis standard, under which the court gives deference to the agency’s interpretation so long as it is reasonable, is applied where the question at issue implicates agency expertise or the determination of fundamental policies within the scope of the agency’s statutory functions. The independent judgment standard, under which the court makes its own interpretation of the statute at issue, is applied where the agency’s specialized knowledge and experience would not be particularly probative on the meaning of the statute. National Bank of Alaska v. State, Department of Revenue, 642 P.2d 811, 815 (Alaska 1982); Kelly v. Zamarello, 486 P.2d 906, 916 (Alaska 1971). We believe that the reasonable basis standard is the appropriate standard of review here because both agency expertise and fundamental policy decisions are involved in the determination of “population,” and because the legislature intended to place the decision in the hands of the department.
Mat-Su urges us to apply the independent judgment standard to the definition of population, arguing that defining who to count, unlike choosing an appropriate methodology for “determining” how many people are in the defined group, involves no agency expertise. We disagree.
The statutes in question direct the department to “determine” population and give the department discretion to choose the data it will use in making its determination, subject only to the requirement that the data be considered reliable. This language acknowledges that determining population is an inexact science, with agency expertise implicated at each step of the determination process.14 Given the expertise involved, as well as the statutory grant of discretion, we evaluate the department’s population determinations applying the reasonable basis standard of review. This conclusion is based on the language and context of the statute, and on subsequent legislative enactments in this area.
First, in the statutory schemes at issue, the legislature provided explicit definitions of difficult terms where it thought them desirable or necessary. For example, revenue-sharing provisions discuss in detail what should be included in the “locally generated revenue” and the “full and true assessed property value” used to calculate the millage rate component of the equalization formula.15 The tax statutes define such terms as “taxable property” and “full and true value” and provide detailed procedures for making property assessments.16 Given this backdrop, it is reasonable to infer that the legislature deliberately left “population” and “residents” undefined, to allow the department to utilize its expertise *176to decide which population groups can feasibly and fairly be counted.17
Further, the recent actions of the legislature support the interpretation we draw from the statutes and their context. In 1985, the legislature passed HB 72, a comprehensive revision and reorganization of the statutes relating to municipal government.18 This bill makes a few minor technical changes in the wording of the sections on population determination for tax-limitation purposes, but the operative phrase— “population shall be determined by the commissioner based on the latest statistics of the United States Bureau of the Census or on other reliable population data” — remains unchanged.19
The population provision in the revenue-sharing statute was amended by the 1985 legislation to read similarly that population shall be determined by “the latest figures of the United States Bureau of the Census or other population data that in the judgment of the department is reliable.”20 The new legislation still does not define “population” or “resident” for the purposes of these sections, and the meaning of the terms was not discussed in the bill analysis, the committee reports, or any of the committee hearings.
Finally, the legislature must have known of the department’s techniques; the department formerly sent an annual report containing its population certifications for revenue-sharing to each legislator, and in more recent years such reports have been available to legislators on request. Given that the legislature must have been aware of the protracted litigation on this point, the reiteration, without amplification, of the problematic phrases concerning population determination can be construed as reaffirming the grant of wide discretion to the CRA.21
Having concluded that the reasonable basis standard applies to the department’s actions in this case, the question then becomes whether the department’s methods for counting population are rational in light of the purposes of the revenue-sharing and tax-limitation statutes. ■
The reasonable basis standard permits the court to consider factors of agency expertise, policy, and efficiency in reviewing discretionary decisions. It is similar to the standard of “unreasonable, arbitrary, and capricious action” under which actions committed to agency discretion are traditionally reviewed when they are subjected to review at all. The reasonable basis standard is appropriate for determining whether the agency deci*177sion has been undertaken “in the manner required by law.”
Jager v. State, 537 P.2d 1100, 1107-08 (Alaska 1975) (footnotes omitted).
With the reasonable basis standard for guidance, we now turn to an examination of the specific methods the CRA utilized in determining population for revenue-sharing and tax-limitation purposes.
2) The 51% Rule for Revenue-sharing
The issue here is whether it was rational for the CRA to use the 51% rule to develop 1983 population figures for revenue-sharing purposes. We believe that it was.
According to the statement of purpose in the 1980 revenue-sharing bill, the objective of the statute is to provide “for more equitable allocation of financial resources among municipalities” and to “assure that no municipality suffers impoverishment of necessary public services, relative to other municipalities, because of the chance location of taxable wealth in the state.”22 Population figures are utilized in the equalization formula as an indication of the relative amount of necessary public services each municipality provides. See AS 29.88.-010. In other words, population figures provide a way to compare one borough to the next so that revenue-sharing funds can be fairly allocated between them.
In addition to developing population data that indicated the relative impact on services provided by each borough, the CRA also wanted to supply the same population figures to various federal programs that provide funds to Alaska. However, since the U.S. Census Bureau would not accept any approach having the potential to count people more than once, the CRA, in order to develop a set of figures usable for both state and federal purposes, necessarily had to select a method of determining population that avoided any hint of double counting.
In determining population for state revenue-sharing purposes we think it was rational for the CRA to develop figures that would be acceptable to the federal government as well. Although satisfying U.S. Census Bureau standards necessarily required a conservative approach to determining population, we believe that such an approach still adequately satisfies the objective of providing a way to compare one borough with the next. Under these circumstances it was reasonable for the CRA to avoid the extra inconvenience and expense of working up separate figures usable for federal purposes alone. Given that it was reasonable to determine population in a way that satisfies federal standards, we also believe it was reasonable for the CRA to select the 51% rule as the method for determining population. Use of this method is acceptable to the U.S. Census Bureau because it avoids double counting. Furthermore, the 51% rule also does an acceptable job of measuring impact on public services because, given the requirement of counting each person at only one place, it makes sense to count each person at the place where that person spends the most time and presumably consumes the most public services.
In short, we believe that the 1983 revenue-sharing figures based on the 51% rule represent a reasonable exercise of the department’s discretion.
3) The Inclusion of the Average Remote Site Workforce in the North Slope Borough’s Tax-limitation Count
The North Slope Borough is different from the other boroughs because of the year-round presence of a stable remote site workforce23 composed of individuals who *178tend to leave the North Slope during their time off.24 This creates a peculiar situation in terms of population data because the various standard methods of determining population differ sharply in their treatment of remote site workers. For example, because this workforce happens to be composed of different individuals at different times, some population methodologies, such as the 51% rule, fail to take account of its presence in the North Slope and instead count the workers only at other locations where they spend more of their time during the year. In contrast, other standard methods for determining population, such as the four-night rule, will account for much of the North Slope workforce because many of the workers will be at the worksite for at least four nights during census week.
Two questions are raised concerning the North Slope’s unique situation: 1) was it rational for the CRA to include the average remote site workforce in the North Slope Borough’s 1983 population for tax-limitation purposes; 2) if so, was it rational to use employment data to estimate that average population? We believe the answer to both questions is yes.
There is no question that including the average remote site workforce in the North Slope Borough’s population count has the potential for some double counting. For example, if person X usually resides in the Mat-Su Borough he would be counted there, and if X also spends a third of the year filling a job on the North Slope he would be counted towards that as well. We believe that the CRA’s decision to determine borough population in a way that might simultaneously credit the same group of people to more than one borough for tax-limitation purposes is not, standing alone, unreasonable. We base this conclusion on two grounds. First, the requirement that the federal government be provided with conservative population figures based on a method that strictly avoids double counting would already be satisfied by the data developed using the 51% rule under the revenue-sharing statute. Thus the choices of methodology available for tax-limitation purposes need not be limited by existing U.S. Census Bureau policies. Second, we see nothing in the purpose of the tax-limitation statute that would prohibit the CRA from determining population in a way that accounts for the average number of people present in the North Slope Borough on any given day. Under the tax-limitation statute, each borough stands alone: population data is not used to compare one borough to the next, but instead is used simply as a limiting factor on each individual borough’s authority to tax certain oil and gas property. As long as the population figures fairly represent the number of people in each borough it should not matter for purposes of tax-limitation that these figures might add up to be somewhat more than the population of the state as a whole.
Having concluded that some double counting is acceptable, we now turn to the more difficult question of whether the remote site workforce should be counted at all. As to the legislative intent, we note that including the North Slope workforce is consistent with the U.S. Census Bureau’s historical policy of counting certain remote site workers as part of the population of the remote site area, and that this Census Bureau policy was in effect at the time the tax-limitation statute was enacted into law in 1973. Furthermore, counting the remote site workforce in the North Slope’s population was a longstanding practice of the department, apparently acquiesced in by the legislature.25 In fact, if the Governor had not vetoed SB 180 in 1982, the CRA would have been required to include the *179remote site workforce in the North Slope’s population. Thus we conclude that counting this workforce does not appear to directly conflict with the legislature’s intentions.
Mat-Su, however, argues that the remote site workforce should not be included in the North Slope population because the workers in remote sites have little impact on local government services. We first note that the tax-limitation statute nowhere directs the CRA to count only those segments of a borough’s population that consume a certain degree of public services. The CRA, of course, is free to investigate ways of determining population that may reflect more accurately the financial impact on local government. In fact, the.ISER report suggested that individuals be apportioned among various boroughs on the basis of time multiplied by an index number which measures the relative amount of services consumed there. The CRA rejected this complex approach partly because it would be so expensive to implement.
Although the ISER report illustrates that population data might be used to reach very precise conclusions as to impact on local services, we do not believe the legislature intended to force the CRA to follow this route. Rather, the tax-limitation statute simply refers to population, an admittedly crude but nevertheless adequate indication of the demand for local services. Although the CRA has the discretion to develop more precise ways to measure impact, we also believe the agency has the discretion to make use of the statutory assumption that the number of people in a borough is by itself a sufficient figure for tax-limitation purposes.26
In sum, we conclude that the CRA acted within its discretion when it included the average remote site workforce in the North Slope Borough’s 1983 population figures for tax-limitation purposes.
We turn now to the second question of whether it was rational to use employment data to estimate the average workforce population.
The tax-limitation statute directs the CRA to determine population, and we believe it is quite reasonable for the CRA to take a practical approach to this task. Indeed, no one seriously suggests that physical head counting is the only permissible way to determine population. We believe that the use of employment data to ascertain the average population of the North Slope workforce is a practical and sensible approach given the fact that virtually every person present in a remote work-site is there because of a job.
The department takes an equally practical approach to counting people in other boroughs. For example, the population of Anchorage has been calculated in past years with a technique known as the “Housing Unit Method.” Under this method, in 1981 and 1982 Anchorage took the number of its household units, categorized by seven structure types, and multiplied this number by the average household size for each type of structure. It then subtracted the estimated number of vacant units, based on the results of a “windshield survey” or on personal calls at a selected sample of houses. It added the military population of the area, based on figures provided by the military, and the number of people living in group quarters, based on the results of a telephone survey. In 1982, Anchorage also added the estimated number of residents in recreational vehicle parks, based on a survey of every fifth vehicle in those parks which gave permission for on-site surveys. Surely estimating *180the average population of a remote site workcamp on the basis of employment data is no less acceptable than estimating the average population of a major city on the basis of house counts and random surveys.
Furthermore, the employment data the department chose to use for the North Slope Borough is reasonably reliable.27 The figures came from monthly reports which employers are required to submit to the Department of Labor for unemployment tax purposes, showing the number of jobs in the pay period containing the twelfth of each month. Mat-Su complains that this data overstates the number of remote site workers because it might include multiple workers who rotate in and out of a single position during a single pay period. The record, however, indicates that the figures refer to the number of jobs in the pay period and not the number of workers who occupy those jobs. It also makes sense to rely on data collected by an impartial state agency rather than, for example, a special census conducted by the municipality that stands to gain from an increased count. Since the data submitted to the Department of Labor is for tax purposes, employers certainly have no incentive to overcount.
In summary, we conclude that it was rational to include the average remote site workforce in the North Slope’s tax-limitation population, that it was rational to use employment data to estimate the average size of that workforce, and that the data used was reasonably reliable.
4) The Use of Different Methods for Revenue-sharing and Tax-limitation Population Counts
The issue here is whether it was reasonable to determine population for revenue-sharing purposes under one method while determining population for tax-limitation purposes under another method. While we note that justifying the use of two methods presents perhaps the greatest conceptual difficulty in this case, we are nevertheless convinced that the CRA’s actions were rational.
When the legislature uses the same term in two closely related statutes, we will normally presume that the legislature intended that term to mean the same thing in both cases. Here, however, there are three factors that, when considered together, are sufficient to overcome the presumption that “population” must mean the identical thing under both statutes. These factors are 1) the differences between the two statutory schemes, 2) the extensive grant of discretion to the CRA, and 3) the legislative acquiescence in the exercise of that discretion. We believe that under these circumstances it was reasonable for the CRA to use a conservative, federally acceptable method for revenue-sharing purposes and at the same time use an equally rational method for tax-limitation purposes that fairly and equitably accounts for the unique situation created by the remote site workforce on the North Slope.
The first of the factors supporting our conclusion involves the differences between the two statutory schemes. Under the revenue-sharing statute, population is used as a way to compare each borough to the next so that the boroughs can equitably share the available funds. Under this scheme, a relative gain in population to one borough means a corresponding loss in the share of funds for the others. We think this fact justifies the use of a conservative approach, such as one that avoids double counting and thus is acceptable for federal purposes as well. The 51% rule is objective, conservative, and well suited for revenue-sharing purposes. This does not, however, mean that such a conservative ap*181proach would be required or even desirable for other statutory schemes.
Population figures serve a slightly different purpose under the tax-limitation statute. Instead of providing a way to compare the relative needs of the various boroughs, the population figures under the tax-limitation statute provide a means of measuring local impact without reference to the other boroughs. In this sense, since each borough stands alone in terms of population, including the remote site workforce in the North Slope Borough’s population does not have a direct impact on the funds available to any other borough. At most, the other boroughs might feel repercussions from the impact on the state budget resulting from the North Slope Borough’s increased ability to tax certain oil and gas property that otherwise would be taxed by the state. However, as there is no credit or refund from the state to the taxpayer for local oil and gas taxes exceeding the state rate of 20 mills, there is no drain on state funds from municipal rates above that amount.28 Furthermore, even if local rates were held to a lower limit, there is no guarantee that any extra funds the state might receive would be allocated in ways affecting the other boroughs. Therefore, and in direct contrast to the effect under the revenue-sharing statute, any change to the North Slope’s tax-limitation population can have at most only an indirect effect on the other boroughs. We believe this distinction goes a long way towards justifying the department’s use of different methodologies to coincide with the different purposes population figures serve under the two statutes.
Nevertheless, without more, we would be hard pressed to say that population can mean two different things under these two related statutes. Here, however, there is something more. In concluding that the rational basis standard of review applies in this case we have already noted the wide discretion granted to the CRA by the legislature. The statutes in question direct the department to “determine” population and give the department discretion to choose the data it will use in making its determination, subject only to the requirement that the data be reliable. We think the scope of the CRA's discretion is broad enough to permit the development of two different sets of population figures in order to better effectuate the purposes of the two statutes. The conclusion that the CRA is vested with broad discretion in implementing these statutes is bolstered by the language in the 1973 amendment to the tax-limitation statute. It states that if two municipalities should ever tax the same property, then:
the commissioner of community and regional affairs shall apportion the lawful levy and equitably divide these revenues on the basis of need, services performed and other considerations in the public interest.29
In selecting its methods for determining population, we think the CRA likewise has the. discretion to consider the special needs and services performed by the respective boroughs. Since the legislature has given the CRA a generous amount of leeway in implementing these statutory schemes in the public interest, we think the CRA is free to use different population figures as long as these figures are reasonable in relation to the particular statute involved. Since we have already concluded that the figures developed by the CRA are individually acceptable, we see no reason to alter that conclusion simply because those figures are not identical.
Furthermore, in keeping with the public interest, we believe the CRA was entitled to consider the legislature’s special concern for the North Slope as demonstrated by the 1982 passage of SB 180. This bill, had it not been vetoed by the Governor, would have required the CRA to include the remote site workforce in the North Slope’s population under both statutes. In our view, the CRA has developed a workable compromise by including the controversial *182remote site workforce for the purposes of one statute but not the other. Although the North Slope may not receive as much funding under the revenue-sharing plan as perhaps it should, the North Slope is nevertheless allowed to make up the loss through higher local taxation which neatly avoids any direct financial impact on the other boroughs.30
In short, we believe the approach taken by the CRA was within its discretionary powers. We are even further persuaded when we consider that the legislature has not altered the scope of the CRA’s authority, although the legislature is certainly free to do so if it chooses. Until that time, however, we see no reason to disturb the existing statutory schemes, vesting as they do broad discretion in the CRA. Therefore we conclude that the use of two population figures for 1983 was reasonable.
B) Administrative Regulations
The 1983 revenue-sharing population counts were determined pursuant to a set of regulations properly issued by the department. However, Mat-Su asserts that the department’s interpretation of population for tax-limitation purposes should have been adopted as a regulation under the procedures outlined in AS 44.62.180-290. Although the tax-limitation statutes do not require that the department promulgate regulations, the department’s interpretation of “population” falls within the statutory definition of a regulation. Therefore, we conclude that the department should have promulgated its approach to the 1983 tax-limitation population counts as an administrative regulation.
Alaska Statute 44.62.640(a)(3) defines a “regulation” as
every rule, regulation, order, or standard of general application or the amendment, supplement or revision of a rule, regulation, order or standard adopted by a state agency to implement, interpret, or make specific the law enforced or administered by it, or to govern its procedure, except one which relates only to the internal management of a state agency; ... “regulation” includes “manuals,” “policies,” “instructions," “guides to enforcement,” “interpretative bulletins,” “interpretations,” and the like, which have the effect of rules, orders, regulations or standards of general application, and this and similar phraseology shall not be used to avoid or circumvent this chapter; whether a regulation, regardless of name, is covered by this chapter depends in part on whether it affects the public or is used by the agency in dealing with the public.
The department’s interpretation of “population” is a regulation because it makes specific a law which the agency administers and because it is used by the agency in dealing with a segment of the public. See Kenai Peninsula Fisherman’s Cooperative Association v. State, 628 P.2d 897 (Alaska 1981) (holding that policies which make specific the broad discretionary power of the Board of Fisheries and serve as the basis for decisions which affect fishermen should have been promulgated as regulations).
The regulation issue was not addressed in the parties’ motions for summary judgment below and thus was not addressed by the superior court’s decision. All of the briefs filed before this court do raise the issue, however, and none of the parties suggests that consideration of the matter be foreclosed because it was not raised below. Where consideration of an issue that has not been properly raised in the trial court is critical to a proper and just decision, we will not decline to consider it, especially after all the parties have had an opportunity to brief it. Vest v. First National Bank of Fairbanks, 659 P.2d 1233, 1234 n. 2, reh. granted 670 P.2d 707 (Alaska 1983). Therefore we modify the *183holding below with an order that the department issue appropriate regulations as soon as possible.
The fact that the department’s interpretation of “population” for tax-limitation purposes should have been properly adopted as a regulation does not mandate a redoing of the 1983 population counts. This case comes before us in the posture of a declaratory judgment. The purpose of a declaratory judgment is to clarify future relations, “to serve some practical end in quieting or stabilizing an uncertain or disputed jural relation.” American Building and Loan Association, Inc. v. State, 376 P.2d 370, 373 (Alaska 1962). Alaska Statute 22.10.020(b) on declaratory judgments provides that:
In case of an actual controversy in the state, the superior court, upon the filing of an appropriate pleading, may declare the rights and legal relations of an interested party seeking the declaration, whether or not further relief is or could be sought. The declaration has the force and effect of a final judgment or decree and is reviewable as such. Further necessary or proper relief based on a declaratory judgment or decree may be granted, after reasonable notice and hearing, against an adverse party whose rights have been determined by the judgment.
Mat-Su’s motion for summary judgment is consistent with the statute; it requests no specific remedies other than “a declaratory judgment establishing the limits of the law and the duties of the parties,” and it con-‘ cedes that “(fjurther remedies would require further consideration by the parties and the court.”31
The department’s 1983 tax-limitation figures are invalid because they were premised on an interpretation which was not properly adopted as a regulation. Therefore there can be no future reliance on this interpretation until the valid procedures are observed. Kenai Peninsula Fisherman’s Cooperative Assoc v. State, 628 P.2d at 906.
In the event that the department adopts regulations which would have produced a lower tax-limitation figure for the borough if they had been applied in 1983, additional proceedings for retroactive relief would not be appropriate. As Mat-Su conceded at oral argument, it would work an unfair hardship on the North Slope Borough to ask it to refund to the state treasury a portion of tax revenues which has already been spent. Furthermore, because there is no way to guess how the legislature might have chosen to appropriate additional revenues, none of the plaintiffs can show that they would have benefitted if a lower borough tax rate had enabled the state to collect more money in oil and gas taxes.
Therefore, our holding on this issue is limited to a declaration regarding the necessity for validly adopted regulations before the Department of Community and Regional Affairs makes future population determinations for tax-limitation purposes.32
*184IV. CONCLUSION
Applying the reasonable basis standard of review, we AFFIRM the superior court’s holding that the Department of Community and Regional Affairs’ 1983 population determinations were reasonable in every respect, modified by an ORDER that the department make all future population determinations for tax-limitation purposes pursuant to validly adopted regulations.
. Effective January 1, 1986, many of the statutes discussed here have been renumbered and in some instances slightly reworded for clarity. Unless otherwise stated, the citations throughout the opinion refer to the statutes in effect at the time the suit was initiated.
. AS 29.88.010 provides:
State equalization of tax resources for local government services, (a) During each fiscal year the department shall compute an equalization entitlement for local government services provided by a taxing unit.
(b) The equalization entitlement computed for a taxing unit is based on the population, relative ability to generate revenue, and local tax burden of the taxing unit and is determined by the application of the formula
Entitlement = P x R
where P = population, and
R = millage rate equivalent, determined by dividing the sum of the locally generated revenue of the taxing unit by one-tenth of one percent (0.1) of the full and true value of assessed property of the taxing unit determined under AS 29.88.020(d); however, the property value used under this subsection may not be less than 15 percent of the statewide average per capita full and true assessed property value.
(c) For purposes of this section, locally generated revenue
(1) includes
(A) the actual revenue derived from the levy and collection of local taxes in the taxing unit for local government services during the preceding fiscal year of the taxing unit;
(B) motor vehicle payments received by the municipality during the preceding fiscal year under AS 28.10.431;
(C) revenue from fees, rentals, leases, penalties, licenses or permits received during the preceding fiscal year by the municipality for a function or service over which it has control, including revenues derived from parks and recreation services, mass transit, offstreet parking, and garbage and solid waste disposal services;
(D) special assessments received during the preceding fiscal year; and
(E) payments received by a municipality from a utility which are in place of taxes levied and collected by the municipality;
(2) excludes
(A) revenue derived from the levy and collection of municipal taxes and appropriated for the operating expenses and debt service of utilities;
(B) revenue from interest earned on investments and from the sale and lease of land or equipment; and
(C) all other revenue from whatever service derived.
. There is, however, a separate provision to guide population determinations under Chapter 89, which provides supplemental state aid to municipalities for specific services such as roads and hospitals on the basis of the magnitude of service provided. Only one category of service is measured on the basis of population; AS 29.89.040 authorizes aid to volunteer fire departments serving an area not in an organized borough or city at the rate of "$10 per capita for the population served by the department, as determined by the state fire marshal.” AS 29.-89.060 states that:
For purposes of AS 29.89.010-29.89.100, population shall be determined by the latest figures of the United States Bureau of the Census or other reliable population data, including but not limited to public school enrollment figures, public utility connection, registered voters or certified employment payrolls.
None of the parties to this litigation has chai-' lenged the population determinations made pursuant to AS 29.89.060.
. AS 29.53.045 provides:
Tax on oil and gas production and pipeline property, (a) A municipality may levy and collect taxes on taxable property taxable under AS 43.56 only by using one of the methods set out in (b) or (c) of this section.
(b)A municipality may levy and collect a tax on the full and true value of taxable property taxable under AS 43.56 as valued by the Department of Revenue at a rate not to exceed that which produces an amount of revenue from the total municipal property tax equivalent to $1,500 a year for each person residing within its boundaries.
(c) A municipality may levy and collect a tax on the full and true value of that portion of taxable property taxable under AS 43.56 as assessed by the Department of Revenue which value, when combined with the value of property otherwise taxable by the municipality, does not exceed the product of 225 per cent of the average per capita assessed full and true value of property in the state multiplied by the number of residents of the taxing municipality. For purposes of this subsection the average per capita assessed full and true value of property in the state shall be calculated without regard to the assessed value of taxable property under AS 43.58.
(d) By February 1 of each assessment year a taxing municipality must inform the Department of Revenue which method of taxation the municipality will use.
(e) For purposes of this section, population shall be determined by the commissioner of community and regional affairs based on the latest statistics of the United States Bureau of the Census or on other reliable population data, and shall advise each municipality of its population as so determined by January 15 of each year.
. AS 29.53.050 provides:
Tax limitation, (a) No municipality may levy and tax for any purpose in excess of three per cent of the assessed valuation of property within the municipality in any one year.
(b) No municipality, or combination of municipalities occupying the same geographical area, in whole or in part, may levy taxes (1) which will result in tax revenues from all sources exceeding $1,000 a year for each person residing within their boundaries or (2) upon values which, when combined with the value of property otherwise taxable by the municipality, exceed the product of 225 per cent of the average per capita assessed full and true value of property in the state multiplied by the number of residents of the taxing municipality. If two or more municipalities occupying the same geographical area, in whole or in part, attempt to levy a tax (1) the combined levy of which would result in tax revenues from all sources exceeding $1,000 a year for each person residing within their boundaries or (2) upon values which, when combined with the value of property otherwise taxable by the municipality, exceed the product of 225 per cent of the average per capita assessed full and true value of property in the state multiplied by the number of residents of the taxing municipality, the commissioner of community and regional affairs shall apportion the lawful levy and equitably divide these revenues on the basis of need, services performed and other considerations in the public interest. For the purpose of this subsection, population shall be determined by the commissioner of community and regional affairs based on the latest statistics of the United States Bureau of the Census or on other reliable population data. For purposes of this subsection the average per capita assessed full and true value of property in the state shall be calculated without regard to the assessed value of taxable property under AS 43.58.
(Emphasis added.) The discrepancy between the $1,000 per capita limit here and the $1,500 per capita limit in AS 29.53.045 was an oversight; both statutes originally contained a $1,000 limit, and the per capita limit here was recently raised to $1,500 to make the two statutes parallel again. AS 29.45.090 (effective January 1, 1986)*.
. Neither the revenue-sharing provisions nor the tax-limitation provisions explicitly require the implementation of regulations to govern population determinations. See AS 29.88.040, AS 43.-58.160 and AS 43.58.190(1).
. The relevant portions of 19 AAC 30.041 provide:
(a) The director will annually determine the population as of July 1 for each city and borough within the State.... The director will base his determination upon the latest official estimate, survey, or census of the United States Bureau of the Census.
(b) The director will, in his discretion, grant an adjustment of the official estimate, survey or census of the United States Bureau of the Census upon his own motion upon the request of another state agency which has authority to make an estimate, survey or census of population, or upon the request of the municipality. If a municipality requests an adjustment under this subsection, the municipality must substantiate the adjustment requested by
(1) a census conducted in accordance with (f) of this section; or
(2) an estimate of the population of the municipality as of July 1 of the entitlement year based on information reported by the most recent federal or local census.
(c)The municipality may include in the census or estimate only those persons who would be considered residents of the municipality by the United States Bureau of the Census.
[[Image here]]
(f) For the purposes of (b)(1) of this section, ... [t]he census must be conducted by the United States Bureau of the Census or by the municipality in a manner satisfactory to the division. The division will validate the census in accordance with standard census definitions and procedures specified by the division or by the United States Bureau of the Census....
. 2 Senate Journal 1788-92 (2d Sess. 1982).
. This is only a slight departure from the pre-1980 U.S. Census method. The U.S. Census Bureau began with the concept of "usual residence,” while the department applied "the 51% rule” to determine base populations. And, while the U.S. Census Bureau applied “the four-night rule” to all workers in group living quarters prior to 1980, the department applied "the four-night rule” only to remote site workers on the North Slope.
. The department estimated that of the 1,850 people added to the North Slope Borough’s count, 307 came from out-of-state while 1,543 had to be subtracted from other municipalities’ counts.
. For the department’s consideration, state demographer David Swanson prepared three sets of 1982 population counts determined under three different methods: 1) "the 1980 U.S. Census method” which counted remote site workers in Alaska at their “home" addresses; 2) "the 51% rule” embodied in the 1981 regulations; and 3) "the 51% rule” as adjusted by the application of "the four-night rule” to the North Slope’s remote site workcamps, pursuant to the emergency regulations.
The following table shows the ramifications of each method for the parties to this suit:
July 1, 1982 Population Estimates
[[Image here]]
The figures prepared in accordance with the emergency regulations were the ones which the department certified.
. The 1981 regulations, which embodied the 51% rule, remained in effect after the court struck down the emergency regulations. See 19 AAC 30.150(10).
. The court granted leave to intervene as plaintiff to Robert J. LeMoine, a resident of the Matanuska-Susitna Borough who works on the North Slope. Because none of the parties has challenged the superior court’s decision that LeMoine has standing as a plaintiff "in the public interest,” we do not reach this issue.
. In the context of reapportionment, we have recognized that the definition of population and the methodology used to count it can be inextricably intertwined. See Egan v. Hammond, 502 P.2d 856 (Alaska 1972).
. See AS 29.88.010 and AS 29.88.020.
. See AS 43.56.210, AS 43.56.060 and AS 29.53.-060.
. See State, Department of Labor v. University of Alaska, 664 P.2d 575, 579 (Alaska 1983) ("where a statute delegating power to an administrative agency does not expressly define the extent of that power, it may be implied from the general policy and purposes underlying the legislative enactment”); North Slope Borough v. LeResche, 581 P.2d 1112, 1115 (Alaska 1978) (concluding, on the basis of legislative history and statutory context, that the legislature intended to give the Department of Natural Resources discretion to reject borough land applications where the land selections would be contrary to the state’s best interests).
. Ch. 74, SLA 1985 (effective January 1, 1986).
. Ch. 74, § 12, SLA 1985 at 115-16. (AS 29.53.-045(e) is renumbered as AS 29.45.080(e), and the sentence about population in AS 29.53.-050(b) is renumbered as AS 29.45.090(d)).
The House adopted the following letter of intent from the Committee on Community and Regional Affairs:
It is not the intent of the House Community and Regional Affairs Committee in adopting AS 29.53.045 as the renumbered section 29.-45.080 in CSHB 72 (C & RA) to alter the substance or effect of that provision.
Legislative Reporting Service (April 1, 1985) at 549.
. Ch. 74, § 16, SLA 1985 at 157. The specific direction to count the military in calculating the equalization entitlement is eliminated (compare AS 29.88.015 to renumbered AS 29.60.020). The list of possible sources of reliable data for calculating population for per capita allocations has also been dropped (compare AS 29.89.060 to renumbered AS 29.60.150).
. Subsequent legislation declaring the intent of a previous enactment is entitled to great weight, and even where subsequent amendments do not explicitly purport to clarify earlier enactments, they may still be probative. Chugach Natives, Inc. v. Doyon, Ltd., 588 F.2d 723, 731 (9th Cir.1978).
. Ch. 155, § 1, SLA 1980.
. The Department of Labor compared monthly employment figures to average annual figures for each census area in the state in order to chart the monthly deviation from the average. The results show the North Slope’s worker population to be relatively steady year-round. In 1983, its monthly deviation ranged from a January population that was 89.1% of the average to a September population that was 108.7% of the average. Only six of Alaska’s twenty-nine census divisions exhibited less seasonal variation in 1983.
. At this time, the only other municipality with permanent remote worksites is the Kenai Peninsula Borough, which reports that most of the workers in these camps are permanent residents of the borough.
. Cf. National Bank of Alaska v. State, Dept. of Revenue, 642 P.2d 811, 813 (Alaska 1982) (even under independent judgment standard, court should give some weight to what agency has done, especially where agency interpretation is longstanding).
. Even assuming arguendo that remote site workers cannot be counted unless there has been some showing of impact on local services, we still would not reverse on that ground because there is evidence in the record that the North Slope Borough provided various services to the remote work camps during the period in question. Furthermore, it is not this court’s function to evaluate the strength of evidence presented to administrative agencies. Oceanview Homeowners Ass’n v. Quadrant Const., 680 P.2d 793, 798 (Alaska 1984). Therefore we need not decide if it would be an abuse of discretion to include an identifiable group in a borough’s population count when there was clearly no impact on the borough from that group.
. In the reapportionment context, this court has upheld reliance on less precise data and techniques. See Kentopp v. Anchorage, 652 P.2d 453, 459-60 (Alaska 1982) (population data found to be reliable, even though it omitted residents in group quarters and non-traditional housing and was based in part on a 2% sampling); Grok v. Egan, 526 P.2d 863, 868 (Alaska 1974) (holding that it was rational for the Reapportionment Advisory Board to base a 1973 count on 1970 data).
. See AS 43.56.010.
. AS 29.53.050(b).
. Using this technique for the North Slope alone is also reasonable because no other borough is in the same situation of providing for a substantial year-round workforce that slips through the cracks, so to speak, in the revenue-sharing context. If another borough should come to face the same circumstances, this same method should be available to it as well.
. Plaintiff-intervenor LeMoine originally requested retroactive relief in the form of a refund to the state treasuries of any taxes collected by the North Slope Borough which should have been collected as state taxes. However one counsel speaking for all the plaintiffs indicated during oral argument that LeMoine now acquiesces in the four municipalities’ request for prospective relief only.
. An analogous situation was presented in Alaska Community Colleges’ Federation of Teachers v. University of Alaska, 677 P.2d 886 (Alaska 1984). After holding that a meeting at which the Board of Regents decided to merge two colleges was conducted in violation of the state Open Meetings Act (OMA), this court addressed the practical problems of choosing an appropriate remedy:
We offer the following as a guide to the superior court if it finds it must determine whether the UAJ merger should be voided. First, reversal of the consolidation will necessarily involve a waste of public resources. It should not be ordered if open and substantial reconsideration has been achieved, or may be achieved at a third meeting. We do not think costly backtracking directly serves the OMA’s public interest goals where less expensive measures are adequate and available.
However, this case may well present a circumstance where the Board of Regents could not honestly reconsider UAJ consolidation be*184cause of the time and resources the University system had already invested in the merger. If the superior court arrives at such a finding, we conclude that the University must show that the prejudice which would be incurred by the public (if the merger were voided) outweighs the remedial priorities of the OMA. If the superior court determines that the damage to the public good outweighs the benefits derived from the dissolution of the UAJ, it may conclude that no further coercive remedy is appropriate.
Id. at 892.
See also Hartman v. City and County of Denver, 165 Colo. 565, 440 P.2d 778 (1968) (declaratory judgment that municipal ordinance setting forth reapportionment scheme is invalid, but upholding election held under the unlawful scheme "to avert serious disruption of the Denver governmental functions").