Logosm.gif (1927 bytes)
navlinks.gif (4688 bytes)
Hruler04.gif (5511 bytes)

Back to Tax Revision Commission main pageBack to research paper list

Triennial Assessment of Real Property in the District of Columbia
John H. Bowman
January 1998

Home

Bibliography

Calendar

Columns
Dorothy Brizill
Bonnie Cain
Jim Dougherty
Gary Imhoff
Phil Mendelson
Mark David Richards
Sandra Seegars

DCPSWatch

DCWatch Archives
Council Period 12
Council Period 13
Council Period 14

Election 1998
Election 2000
Election 2002

Elections
Election 2004
Election 2006

Government and People
ANC's
Anacostia Waterfront Corporation
Auditor
Boards and Com
BusRegRefCom
Campaign Finance
Chief Financial Officer
Chief Management Officer
City Council
Congress
Control Board
Corporation Counsel
Courts
DC2000
DC Agenda
Elections and Ethics
Fire Department
FOI Officers
Inspector General
Health
Housing and Community Dev.
Human Services
Legislation
Mayor's Office
Mental Health
Motor Vehicles
Neighborhood Action
National Capital Revitalization Corp.
Planning and Econ. Dev.
Planning, Office of
Police Department
Property Management
Public Advocate
Public Libraries
Public Schools
Public Service Commission
Public Works
Regional Mobility Panel
Sports and Entertainment Com.
Taxi Commission
Telephone Directory
University of DC
Water and Sewer Administration
Youth Rehabilitation Services
Zoning Commission

Issues in DC Politics

Budget issues
DC Flag
DC General, PBC
Gun issues
Health issues
Housing initiatives
Mayor’s mansion
Public Benefit Corporation
Regional Mobility
Reservation 13
Tax Rev Comm
Term limits repeal
Voting rights, statehood
Williams’s Fundraising Scandals

Links

Organizations
Appleseed Center
Cardozo Shaw Neigh.Assoc.
Committee of 100
Fed of Citizens Assocs
League of Women Voters
Parents United
Shaw Coalition

Photos

Search

What Is DCWatch?

themail archives

Preliminary Report on
Triennial Assessment of Real Property in the District of Columbia
by John H. Bowman, Ph.D.
Draft report prepared for the District of Columbia Tax Revision Commission January 1998


    Table of Contents    Rationale for Triennial Assessment

Effects on a Single Triennial Group
Steady Increase in Market Value
Faster Increase in Market Value
Declining Market Values
Summary of Assessment-Level Effects

Effects with Staggered Valuation Across Multiple Groups
Steady Growth of Market Value, Phase-in of Changes
Steady Growth of Market Value, No Phase-in of Changes
Irregular Changes in Market Value, Phase-in of Changes
Irregular Changes in Market Value, No Phase-in of Changes
Summary of Effects

Recommendations
Recommendation 1
Recommendation 2

Ending the Triennial System


Tables

Table 1 Hypothetical Examples of Effects of Triennial Assessment with Changes Phased In

Table 2 Triennial Assessment with Phase-in - Hypothetical Example of Effects on Assessment Level and Uniformity Using Homes with Base Period Market Value of $100,000 and Experiencing Annual Value Changes at the Actual Annual Average Percentages Rate for Aggregate Residential Assessed Values for 1986-96, by Triennial Assessment Group

Table 3 Triennial Assessment without Phase-in - Hypothetical Example of Effects on Assessment Level and Uniformity Using Homes with Base Period Market Value of $100,000 and Experiencing Annual Value Changes at the Actual Annual Average Percentages Rate for Aggregate Residential Assessed Values for 1986-96, by Triennial Assessment Group

Table 4 Triennial Assessment with Phase-in - Hypothetical Example of   Effects on Assessment Level and Uniformity Using Homes with Base Period Market Value of $100,000 and Experiencing Annual Value Changes at the Actual Annual Percentages Rates for Aggregate Residential Assessed Values for 1986-96, by Triennial Assessment Group

Table 5 Triennial Assessment without Phase-in - Hypothetical Example of Effects on Assessment Level and Uniformity Using Homes with Base Period Market Value of $100,000 and Experiencing Annual Value Changes at the Actual Annual Percentages Rates for Aggregate Residential Assessed Values for 1986-96, by Triennial Assessment Group

Table 6 Moving from Triennial Assessment System with Phase-in of Valuation Changes to an Assessment at Current Market Value, Selected Points in Triennial Cycle, by Triennial Group


Triennial Assessment

Historically, real property in the District of Columbia has been revalued annually, as required by statute [District Code, 47-820]. In 1997, however, legislation was adopted moving the District to a triennial assessment system [District of Columbia 1997d].(1) Under this system, which is to be effective with the 1999 tax year, the city is divided into three areas with essentially equal amounts of taxable property, each to be revalued in turn. Thus, assessments are to be done on a staggered basis, the first area for tax year 1999, the second for tax year 2000, and the third for 2001; after the completion of the first round, the cycle would begin again with the first area being revalued for 2002, and so on. Moreover, the new values are to be phased in via equal increments in each of the three years between revaluations for each triennial group.

Rationale for Triennial Assessment

A February 1997 briefing packet outlines the Real Property Tax Administration's proposed triennial assessment process and some of the reasons for it. Some problems with the District's real property tax system, including relatively poor performance in recent years in attaining high-quality assessments, have been reviewed in an earlier portion of the report on property taxes. They lie behind the move to triennial assessment. The basic consideration noted in the briefing packet was to allow the limited number of assessors more time to do their job properly.

The existing annual valuation cycle is currently unmanageable. It has not allowed for an effective property inspection and data collection process. If the current annual valuation process remains in effect with existing staff levels, each assessor would be assigned 8,500 parcels of real property to value. Under these circumstances, it would be impossible to keep current with changes in the market and factual data for every property located in the District [District of Columbia 1997d, 2-1].

Under the triennial approach, each assessor is responsible each year for valuing about 2,850 real property parcels.

Another argument for moving away from annual revaluation is that most states do not undertake annual revaluation. It is reported that while 30 states have an annual assessment requirement, only six have a true annual appraisal system [District of Columbia 1997d, 3-1, 4-1, 5-1]. The District's briefing packet relies for this information upon a 1992 survey by the International Association of Assessing Officers (IAAO). A similar picture emerges from the Census of Governments.

Statutory language appears to require annual assessment in 28 states and the District of Columbia [U.S. Bureau of the Census 1989, Appendix D]. Although I cannot get the number of annual-valuation states as low as six from this source, the wording in several instances is quite vague, so that the requirement may be different from what it seems. For example, the Census report indicates that in Alabama the "tax assessor has the right and authority to assess real estate annually" [emphasis is mine]; Georgia is said to provide for "the opening of books for return of taxes each year"; and in several states the provision is only for production of an annual tax roll, not necessarily of new values.

From the provisions reported by Census, it is clear that in some states annual assessment does not mean annual physical inspection or complete reappraisal. For example, Nebraska requires "an annual review of the appraised values for the purpose of maintaining and updating the assessment roll" but "a complete reappraisal" is to be undertaken only "when ordered by the Tax Commissioner." Similarly, in Utah "assessors are required to visit each separate district or precinct . . . , annually, including inspection where necessary" [emphasis is mine].

Even where there appears to be a requirement for annual valuation, it is known that values are not always changed annually. One researcher found, in fact, that states with requirements for revaluation less frequently than annually generally had better results - more uniform assessments - than the presumed annual-valuation states [Mikesell 1980]. Apparently it often is not feasible to carry out a complete revaluation annually, and so the requirement is considered unrealistic and is not met. And once the annual requirement is ignored one year, two years, or more, it is not obvious when revaluation actually should occur. States with a stipulated multi-year cycle, however, seem more likely to enforce the requirement, thus assuring that revaluation does, in fact, occur periodically.

Based on the Census report noted above, legally-provided multi-year assessment cycles range from two years to 10. Iowa reports a two -year cycle, and at the other extreme, both Connecticut and Rhode Island have 10-year cycles. The most common assessment-cycle lengths are four years and five years, with six states each. Again making the distinction between revaluation and complete reappraisal, two states with longer assessment cycles (Ohio six years, North Carolina eight years) seem to require complete reappraisal at the stipulated intervals, but also require values to be updated by less costly means in the middle of those cycles.

In short, the District has plenty of company in opting to abandon the once-typical requirement that all real property be reappraised annually. One neighboring state, Maryland, has a triennial system very similar to that now established for the District. Virginia requires revaluation every 2-4 years for independent cities and every 4-6 years for counties; for each type of unit, those with smaller populations are permitted to go longer between revaluations. There, localities may elect to revalue more often than required, and many of the larger ones do so annually; where this is done, however, annual revaluation typically does not mean annual physical inspection. As already noted, available evidence suggests that multi-year cycles can result in comparatively good assessment results.

Thus, the rationale for the staggered triennial assessment system is, in part, the likelihood of improved assessments. Indeed, assessment uniformity in the District probably will improve in the coming years. Nonetheless, the triennial process itself tends to produce assessed values that are lower overall and less uniform than otherwise might be gotten. A basic reason is that assessments always will lag behind market developments.

In the following pages, these tendencies are explored under different circumstances regarding the rate of change in market value and whether or not new values are phased in or implemented immediately. Considered first are the effects within a single triennial group. Subsequently, effects across the three triennial groups are examined. Because there is no experience with the District's triennial system, hypothetical data are used.

Effects on a Single Triennial Group

Steady Increase in Market Value

To illustrate the effects of the triennial assessment system, an example is constructed in which there is assumed to be no assessment error from other sources (table T1). Thus, in the base period, the market value is $100,000; the city valuer's estimate of market value (here termed appraisal value) is $100,000; and the assessed value (the tax base) also is $100,000. Because both the assessed value (AV) and market value (MV) are $100,000, the ratio of assessed value to market value (AV/MV) is 100 percent. AV/MV ratios are the same as the assessment-sales ratios considered in an earlier installment of the property tax report.

The first panel of table T1 uses a 5 percent annual increase in market value. Under the assumption that city valuers are accurate, the appraisal exactly reflects this increase. Thus, both market and appraisal values rise to $105,000 in the first year of the triennial system. However, due to the phase-in of the increase over the three-year cycle, only one-third of the $5,000 increase shows up in assessed value - the actual tax base - each year. By the third year of the cycle, the home finally is valued at its market value in the first year of the cycle. Because of this timing, the ratio of assessed value to market value (AV/MV) falls from 100 percent in the base year to 96.8 percent in the first year of the triennial cycle, and then to 93.7 percent and 90.7 percent, respectively, in the last two years of the first triennial cycle.

The next year - year 4 in the table - it is time for the home to be reappraised once more, for the first time in three years. Meanwhile, market value has risen by 5 percent each year, so the market value in year 4 is $121,551. Again assuming accurate valuation, the appraised value also is $121,551. The difference in appraised values between the first and second triennial valuations is $16,551 (121,551 - 105,000). As before, only one-third of this increase is picked up on the assessment roll each year, so the AV/MV value continues to stay well below 100 percent. In fact, in this example, the assessment level stays just over 90 percent after full implementation of the triennial system - i.e., from year three. Three key assumptions underlie this outcome:

  • the annual percentage increase in market value is constant
  • the value of that percentage increments is five percent
  • the valuer's estimate market value precisely at market value in each triennial valuation

As later examples will show, deviations from these assumptions change the outcome.

Faster Increase in Market Value

One such deviation is a change in the assumed rate of change in market value. The second panel of table T1 shows a constant rate of increase of 10 percent per year. The result is the same as in the first panel (5 percent annual increase), except that the annual dollar increases in market value and assessed value are higher while the level of assessed value to market value falls lower. Combining a 10 percent annual increase in market value with revaluation every third year and increases in assessed value phased in over a three-year period causes the ratio AV/MV to drop to approximately 83 percent by the third year of the initial triennial cycle.

Thus, taken alone, triennial valuation changes phased in over the three-year cycle would lower the average assessment level in a period of rising market values, for the assessments on the tax roll always would lag behind the market. The magnitude of the reduction increases as the rate of increase in market value rises.

Declining Market Value

Drops in market value also have been experienced in the District of Columbia, as well as in other areas, in recent years. The third panel of table T1 shows the effects of a 3 percent annual drop in home value, together with the triennial system. Because estimating new appraised values only once every three years causes assessed values to lag market values - and phasing in the new values over the three-year cycle amplifies the lag - annually falling market value causes assessed value to be above market value. In the example of table T1, with a 3 percent annual fall in market value, the ratio AV/MV rises to a bit over 106 percent by the third year of the initial triennial cycle and remains there.

Summary of Assessment-Level Effects

Using the simplified example of table T1, we see that triennial assessment with phased-in changes in appraised values over the three-year cycle causes the assessment level to depart from market value. If market values rise steadily - or fall steadily - assessed values never will be at market value, even if the initial appraisals by city valuers are on the money. Changes in the value used as the tax base lag changes in the market. This would be true with triennial assessment without a phase-in feature, but the phase-in heightens the departure from market-value assessment. In a period of steadily rising market value, assessed value falls below market value. The higher the rate of increase in market value, the greater the gap between assessed and market values. In a period of steadily falling market value, the results are the same except the departure of assessments from market value is in the opposite direction - i.e., AV/MV becomes greater than 100 percent.

Effects with Staggered Valuation Across Multiple Groups

The examples of table T1 follow changes in only one house - or, alternatively, changes for a representative property in one triennial group. Although helpful as far as it goes, this cannot portray the effects of the staggered (or sequential) triennial cycle adopted for the District. Under that system, approximately one-third of the properties are placed into each of three groups, and the three groups are reappraised in successive years. One group is revalued for 1999, the next for 2000, and the third for 2001, after which the first group is reappraised for 2002, the second for 2003, and so on. Thus, there never is a period in which all properties are to be valued as of the same date. Some degree of assessment non-uniformity is the result. It can be measured by the coefficient of dispersion (COD - explained in an earlier installment of the property tax report). Before getting to COD figures, however, it is important to outline the tables used in this analysis.

Four tables are used to show the uniformity effects of the triennial system, and to amplify upon its implications for assessment level. The first two use a constant annual rate of increase in market value to show the effects of the triennial valuations, with phase-in (table T2) and without (table T3). The annual rate of increase used is 5.3768 percent. This seemingly too-precise figure is the compound annual average increase in residential assessments for the District over the period 1986-96 (reported in last fall's overview section of the property tax report). It is used, in part, because the other pair of tables uses the actual annual percentage increases for that period. By using the exact average, the beginning and ending market value figures are the same in all four tables. As in the first set of tables, the second also shows the effects of triennial assessment with phase-in and without (tables T4 and T5, respectively).

In each of these four tables, the first two columns show the year and the market value for that year (percentage changes are given in footnotes in each table). Starting from $100,000 in the base year, market value rises to $168,830 by the tenth year in all four. In tables T2 and T3, this results from uniform yearly increases of 5.3768 percent, while in tables T4 and T5, the annual growth rates are varied, including two negative changes. The percentage changes in the latter pair of tables range from - 6.7 percent to 15.5 percent, reflecting aggregate actual experience with residential property in the District in the 1986-96 period. The next three columns show assessed values in each year for each of three triennial assessment groups. Under the simplifying assumption of perfect appraisal accuracy, the appraised value for each group is the market value in the year of reappraisal. A further assumption is that there are no changes to the properties that would warrant other valuation changes. As under new District law, the three groups are valued sequentially, one group per year over the three-year cycle. With phase-in, the change in assessed value each year is one-third of the change in appraised (market) value. Without phase-in, the change is posted in the year of reappraisal and then left unchanged for the next two years.

Steady Growth of Market Value, Phase-in of Changes

Group 1 revaluation occurs in year 1. Appraised value rises from $100,000 to $105,377, and the assessed value on the tax roll rises from $100,000 to $101,792. The increase in assessed value - $1,792 - is one-third of the $5,377 change in market value. The same increment is posted on the assessment roll in each of the next two years, so that at the end of the first triennial cycle for group 1, the assessed value has risen to equal market value in the first year of that cycle. Group 2 is revalued for year 2 of the example, when market value is $111,043, so assessed value rises from $100,000 to $103, 681. And for group 3, year 3 marks the launch of the triennial system. By this time, market value is $117,013 - up $17,013 from the base year - so assessed value is increased by $5,671 (one-third of $17,013). The first year of the triennial cycle for each group is shown in bold-face type in all four tables.

There are four columns in the next set of columns. The first three of these give the ratios of the assessed values (from columns 3-5) to market value (column 2), while the fourth column in the set presents the mean ratio for the three triennial groups taken together. In the base year, the ratio of assessed value to market value (AV/MV) is 100 percent, reflecting the assumption of perfect assessment accuracy, made to permit isolation of the effects of triennial valuation. The first year of the triennial approach results in adjustments to assessed values only for group 1, and for that group only one-third of the increase in market value is added to assessed value. Thus, the assessment level (AV/MA) falls for all three groups, but by less for group 1. In year 2, market value again grows by more than 5 percent; because group 3 still has not been revalued, its assessment remains unchanged and the ratio AV/MV drops further. The assessment ratios for groups 1 and 2 also drop more, because the increase in assessed values is less than the increase in market value. In year 3, assessed values are changed for all three triennial groups but, of course, they still lag market developments, so AV/MV values fall more.

For each group, the assessment level approaches 90 percent in year 3 (still a little higher for group 2 until year 4), where it will stay, as long as the percentage increase in market value is the same each year and there is no assessment error to affect the assessment ratio. The mean assessment ratios in the last column of this set show that, for the three groups taken together, the average assessment level falls in steps from 100 percent in the base year to 90 percent by year 4.

The last column presents the coefficient of dispersion (COD - the absolute average deviation of each ratio from the median ratio, expressed as a percentage of the median ratio). Because the example was structured to have no valuation error as such, the CODs measure the degree of assessment non-uniformity attributable solely to the staggered triennial assessment system - i.e., to the practice of estimating value changes only every third year and then phasing in the changes via increases of equal dollar value in each year of the three-year cycle. Because up through the base year assessed values have been estimated annually, the COD value for that year is zero. (Recall that greater degrees of non-uniformity result in higher COD values.) Under the market conditions assumed in table T2, the non-uniformity from triennial assessment features is negligible. After the first triennial cycle is completed (i.e., from year 3 forward), the COD value is below one in this table.

Steady Growth of Market Value, No Phase-in of Changes

Table T3 differs from table T2 in that the latter dispenses with the phase-in of new appraised values, entering the full changes for a triennial group onto the assessment roll in the first year of the triennial cycle. Annual percentage changes in market values remain constant, as in table T2. Eliminating the phase-in introduces more variation across the three triennial groups in the level of assessed values relative market value (AV/MV), which shows up as a higher degree of non-uniformity. The COD values in the last column of the table are five times as high as in table T2 - 3.5 versus 0.7. While even 3.5 is a relatively low COD, this non-uniformity results solely from the staggered triennial assessment cycle. Other sources of assessment error - assumed away in these examples to focus on the effects of the triennial approach - will be present. Those other effects have accounted for CODs greater than zero in the District prior to the introduction of the triennial system.

Comparing tables T2 and T3, the trade-off is between a higher average level of assessment without the phase-in and a lower degree of non-uniformity with the phase-in. The relevant figures are in the last two columns of each table. Abandoning the standard of annual estimation of current market value as the tax base in favor of triennial valuation reduces the average level of assessment (the ratio AV/MV). With the phase-in, the average assessment level falls from 100 percent to 90 percent; without a phase-in, the assessment level is 95 percent. But the phase-in has the effect of smoothing out the differences across triennial assessment groups that result from the staggered revaluation cycle, and this shows up in lower COD values with a phase-in than without.

Irregular Changes in Market Value, Phase-in of Changes

Table T4 relaxes the assumption of uniform annual percentage changes in market value used in the previous two tables. Instead, the pattern of actual annual aggregate market value changes for residential property in the 1986-96 period is used. Except for years 7 and 9, all changes were positive, with values ranging from as small as 3.7 percent to as high as 15.5 percent; the negative figures were - 1.6 percent and - 6.7 percent (see notes to tables T4 and T5 for annual detail). The average over the 10-year period (5.3768 percent) was used in table T2 and T3. As in table T2, value changes in table T4 are phased in.

The primary difference resulting from the pattern of irregular market value changes is the loss of ability to make easy generalizations about the degree of under- or over-assessment relative to market value. In table T2, it was shown that uniform annual increases of roughly 5 percent caused the assessment level for each triennial group in a mature triennial system to be about 90 percent of market value, as shown by AV/MV ratios. Table T1 showed that faster growth resulted in a larger drop in the assessment level, while negative value change caused assessed values to exceed market values. In table T4, the mix of positive and negative market value changes of different sizes causes the value of AV/MV to bounce around over time rather than to move smoothly toward a given value. Moreover, because reappraisal years - and the extent of change in market value between reappraisal years - differ across the three triennial groups, the assessment levels tend to vary more across those groups in any given year. For the mature system - i.e., beginning with year 3 - the assessments levels within group 1 range between 81.2 percent and 100.2 percent. For groups 2 and 3, respectively, the ranges are from 84.5 percent and 103.8 percent and from 79.6 percent to 101.2 percent. For the three groups taken together, the mean assessment level of the mature system varies between 81.8 percent and 101.7 percent.

Because of the specifics of the situation, only group 3 experienced year-to-year decreases in absolute assessed values in the 10-year period considered. Those were in the ninth and tenth years, because group 3 was up for reappraisal in year 9, the year of a significant drop (6.7 percent) in market value. Thus, for the next three years that reduction is phased in for group 3. In the same interval, groups 1 and 2 both receive assessment increases. For the tenth year, the homes in groups 1-3, respectively, are valued at $161,057, $168,173, and $157,846, although all three have the same market value $168,830).

The coefficient of dispersion (COD) values under triennial assessment in table T4 range from 0.8 to 2.7, compared with 0.6 to 1.2 (and 0.7 for the years of the mature system) in table T2, which also includes a phase-in, but features uniform annual percentage changes in market value. Thus, uneven growth in market value results in more variability of assessments - both across triennial groups in a given year, and with each group over time.

Irregular Changes in Market Value, No Phase-in of Changes

The situation portrayed by table T5 is the same as that in table T4, except that new appraised values are entered, in full, onto the tax rolls the first year of the triennial cycle. Thus, properties in each group are valued at their respective market values in the reappraisal year, and those values remain on the tax roll for three years (still assuming no changes in the properties to trigger a revaluation). Compared to the phase-in case, and considering only years 3 through 10 ,lack of a phase-in is associated with a higher average assessment level within each triennial group, generally more variability of assessment levels across triennial groups in a given year, and mixed changes in the uniformity of assessments within each group over time.

Summary of Effects

As noted at the outset, triennial assessment tends to reduce overall assessment levels and to increase variation in assessment levels across property owners. The differences vary, however, with such things as the rate and pattern of annual market value changes and whether or not new values are phased in. To help facilitate comparisons, figures based on tables T2-T4 but not contained in those tables are placed in tabular presentations below.

The first presentation considers assessment level. Particularly for tables T4 and T5, the

variation in AV/MV values within and among triennial groups in tables T4 and T5 makes it difficult to gauge averages by casual inspection. Mean ratios have been computed for the last eight years - i.e., what has been termed the period of the mature triennial system. The mean is more instructive than the median of the overall revenue effects by group and for the sum of all groups, mean values are noted here. As shown in the tabular presentation below, the average assessment level is roughly 5 percentage-points higher without a phase-in that with a phase-in, regardless of whether market values change by uniform percentage increments or follow an irregular path. (Recall that the magnitude of the difference is sensitive to the average rate of change.) Irregular change causes significant variation across groups in terms of the eight-year average assessment level. This was seen earlier via the COD values in tables T2-T5.

Mean AV/MV, years 3-10, for:

Group 1 Group 2 Group 3
Uniform annual market value change, phase-in (T2) 90.2% 90.4% 90.3%
Uniform annual market value change, no phase-in (T3) 95.0% 94.4% 95.6%
Irregular annual market value change, phase-in (T4) 89.2% 93.4% 90.2%
Irregular annual market value change, no phase-in (T5) 93.9% 98.0% 95.1%

While the mean assessment level figures are instructive, they mask year-to-year variation. Coefficients of dispersion of the AV/MV ratios for years three through 10 (calculated with respect to the median ratio) have been calculated to provide some insight into this aspect of the effects of the features of triennial assessment. They are shown in the second tabular box, below. Clearly, the least fluctuation results from uniform annual percentage changes in market value, picked up accurately by the valuers: COD values are less for tables T2 and T3 than for their counterparts in table T4 and T5 - i.e, comparing T2 to T4, and T3 to T5. Another conclusion is that the phase-in feature serves to reduce variation in year-to-year assessment levels within a triennial group. This clearly is the case when value change is uniform (comparing table T2 to T3), but is less clear when market value change is irregular (comparing table T4 to T5).

CODs of AV/MV ratios, years 3-10, for:

Group 1 Group 2 Group 3
Uniform annual market value change, phase-in (T2) 0.08 0.29 0.06
Uniform annual market value change, no phase-in (T3) 3.9 3.2 3.3
Irregular annual market value change, phase-in (T4) 4.7 5.6 6.2
Irregular annual market value change, no phase-in (T5) 6.3 3.7 5.4

In comparing the four cases presented in the four tables, the pattern of non-uniformity revealed by CODs for assessment levels with triennial groups over time is essentially the same as that revealed by the CODs for assessment levels across groups within a given year: There is more non-uniformity due to triennial assessment when market value growth is irregular, but the phase-in feature tends to reduce the measured non-uniformity.

In short, triennial assessment reduces the level of assessment, but there is a trade-off between assessment-level reduction and uniformity. Specifically, we conclude that:

  • staggered triennial assessment reduces the overall level of assessment
  • phasing in value changes amplifies the departure of assessments from market values
  • staggered triennial assessment adds some non-uniformity to assessment outcomes
  • the additional non-uniformity is slight when market value changes follow a smooth path
  • phasing in value changes generally reduces the non-uniformity from the triennial approach

The magnitude of these effects depends upon both the average rate of change in market values over time, and the particular pattern of year-to-year changes.

Recommendations

The District adopted the triennial valuation system in large part due to the lack of resources to reappraise each property every year. In not making annual appraisals, the District joins a large number of states. However, annual revaluation need not mean annual reappraisal in the sense of verifying the data on each property every year, especially via on-site inspection. 

Recommendation 1. Ultimately, the District should move to greater reliance upon a computer-assisted mass appraisal system to permit annual changes in assessed values, where warranted. Annual physical inspection is not recommended or envisioned.

Although the triennial system has just been adopted and is barely off the ground, the recommendation is that it not be retained. A principal reason is the effect on the average level of assessment - i.e., of the tax base. Any reduction in the base implies either a reduction in the level of services, or an increase in the nominal tax rates. Growth averaging roughly the same as in the 1986-96 period (i.e., a little above five percent annually) would result in nearly a 10 percent reduction in the average assessment level. The District's history is one of relatively stable nominal rates. Continuation of this would translate into generally reduced revenue levels, assuming a continued general upward movement of market values.

Another consideration underlying this recommendation is that triennial assessment adds to assessment non-uniformity, particularly when market value changes do not occur in uniform percentage increments. While this effect can be diminished through the phasing in of valuation changes, the phase-in also substantially increases the reduction of the revenue base. Assuming a general upward trend in property prices, triennial valuation causes assessed value to fall below market value in at least two of the three years of the cycle; with a phase-in feature, it assessed value always will be below market value.

While the District had sound reason to seek a change in its assessment process, it is a mistake to equate annual valuation with annual inspection of each property. As a good data base of the relevant attributes of each property is developed and maintained, it will be possible to use this information to generate estimated values of all properties more frequently. It was noted above that several states already draw a distinction between valuation and physical inspection. The latter is particularly onerous, if the requirement is to be meaningful. It also is not essential, when modern computer capabilities are used to develop and maintain the information on real property needed to generate estimated values without physical inspection.

Recommendation 2. If triennial assessment is maintained, consideration should be given to eliminating the phase-in.

The primary reason for this is to avoid the adverse effect on the level of the tax base and, indirectly, on revenues. An offsetting consideration, of course, is that the phase-in tempers the non-uniformity effects of staggered triennial valuation. Also, without a phase-in, the one-year changes in assessed value could at times be large, and larger increases are more unpopular with the public. This was one of the considerations behind the first recommendation, to make smaller annual adjustments in assessed values relying more heavily on CAMA.


1. An up-to-date copy of the code was not available as late as October, but the legislation contained in this briefing packet [tab 12] reportedly was adopted in all essential details. The bill, introduced by David A. Clarke, chairman of the District Council, was to be cited as the Real Property Assessment Process Amendment Act of 1997. The December 1997 issue of Commerce Clearing House's State Tax SourceDocs, a CD-ROM updated monthly, includes the new provisions.


Table T1 Hypothetical Examples of Effects of Triennial Assessment with Changes Phased In

Year Market Value Appraisal Change Assessed Value AV/MV
Base $100,000 $100,000   $100,000 100.0 %
(Case 1 - Values Rise by 5% Per Year)
1 105,000 105,000 5,000 101,667 96.8
2 110,250     103,333 93.7
3 115,763     105,000 90.7
4 121,551 121,551 16,551 110,517 90.9
5 127,628     116,034 90.9
6 134,010     121,551 90.7
7 140,710 140,710 19,159 127,937 90.9
(Case 2 - Values Rise by 10% Per Year)
1 110,000 110,000 10,000 103,333 93.9
2 121,000     106,667 88.2
3 133,100     110,000 82.6
4 146,410 146,410 36,410 121,137 83.4
5 161,051     134,273 83.4
6 177,156     146,410 82.6
7 194,872 194,872 48,462 162,564 83.4
(Case 3 - Values Decline by 3% Per Year)
1 97,000 97,000 -3,000 99,000 102.1
2 94,090     98,000 104.2
3 91,267     97,000 106.3
4 88,529 88,529 -8,471 94,176 106.4
5 85,873     91,353 106.4
6 83,297     88,529 106.3
7 80,798 80,798 -7,731 85,952 106.4
Source: Author's calculations; see text. Rounding may cause slight discrepancies in details.

 

Table T2 Triennial Assessment with Phase-in - Hypothetical Example of Effects on Assessment Level and Uniformity Using Homes with Base Period Market Value of $100,000 and Experiencing Annual Value Changes at the Actual Annual Average Percentages Rate for Aggregate Residential Assessed Values for 1986-96, by Triennial Assessment Group

Year Market Value* 
(MV)
Assessed Value (AV) Ratio: AV/MV (Percent) Coefficient 
of 
Dispersion
Group 1 Group 2 Group 3 Group 1 Group 2 Group 3 Mean
Base $ 100,000 $ 100,000 $ 100000 $ 100,000 100.0 100.0 100.0 100.0 0.0
1 105,377 101,792 100,000 100,000 96.6 94.9 94.9 95.5 0.6
2 111,043 103,585 103,681 100,000 93.3 93.4 90.1 92.2 1.2
3 117,013 105,377 107,362 105,671 90.1 91.8 90.3 90.7 0.6
4 123,305 111,353 111,043 111,342 90.3 90.1 90.3 90.2 0.7
5 129,935 117,329 117,340 117,013 90.3 90.3 90.1 90.2 0.7
6 136,921 123,305 123,638 123,649 90.1 90.3 90.3 90.2 0.7
7 144,283 130,298 129,935 130,285 90.3 90.1 90.3 90.2 0.7
8 152,041 137,290 137,304 136,921 90.3 90.3 90.1 90.2 0.7
9 160,216 144,283 144,672 144,686 90.1 90.3 90.3 90.2 0.7
10 168,830 152,465 152,041 152,451 90.3 90.1 90.3 90.2 0.7
* Market value increases 5.3768% each year after the base year. 

Source: Author's calculations.

 

Table T3 Triennial Assessment without Phase-in - Hypothetical Example of Effects on Assessment Level and Uniformity Using Homes with Base Period Market Value of $100,000 and Experiencing Annual Value Changes at the Actual Annual Average Percentages Rate for Aggregate Residential Assessed Values for 1986-96, by Triennial Assessment Group

Year Market Value* 
(MV)
Assessed Value (AV) Ratio: AV/MV (Percent) Coefficient 
of 
Dispersion
Group 1 Group 2 Group 3 Group 1 Group 2 Group 3 Mean
Base $ 100,000 $ 100,000 $ 100000 $ 100,000 100.0 100.0 100.0 100.0 0.0
1 105,377 105,377 100,000 100,000 100.0 94.9 94.9 96.6 1.8
2 111,043 105,377 111,043 100,000 94.9 100.0 90.1 95.0 3.5
3 117,013 105,377 111,043 117,013 90.1 94.9 100.0 95.0 3.5
4 123,305 123,305 111,043 117,013 100.0 90.1 94.9 95.0 3.5
5 129,935 123,305 129,935 117,013 94.9 100.0 90.1 95.0 3.5
6 136,921 123,305 129,935 136,921 90.1 94.9 100.0 95.0 3.5
7 144,283 144,283 129,935 136,921 100.0 90.1 94.9 95.0 3.5
8 152,041 144,283 152,041 136,921 94.9 100.0 90.1 95.0 3.5
9 160,216 144,283 152,041 160,216 90.1 94.9 100.0 95.0 3.5
10 168,830 168,830 152,041 160,216 100.0 90.1 94.9 95.0 3.5
* Market value increases 5.3768% each year after the base year. 

Source: Author's calculations.

 

Table T4 Triennial Assessment with Phase-in - Hypothetical Example of Effects on Assessment Level and Uniformity Using Homes with Base Period Market Value of $100,000 and Experiencing Annual Value Changes at the Actual Annual Percentages Rates for Aggregate Residential Assessed Values for 1986-96, by Triennial Assessment Group

Year Market Value* 
(MV)
Assessed Value (AV) Ratio: AV/MV (Percent) Coefficient 
of 
Dispersion
Group 1 Group 2 Group 3 Group 1 Group 2 Group 3 Mean
Base $ 100,000 $ 100,000 $ 100000 $ 100,000 100.0 100.0 100.0 100.0 0.0
1 107,300 102,433 100,000 100,000 95.5 93.2 93.2 94.0 0.8
2 116,957 104,867 105,652 100,000 89.7 90.3 85.5 88.5 1.8
3 121,284 107,300 111,305 107,095 88.5 91.8 88.3 89.5 1.3
4 131,957 115,519 116,957 114,189 87.5 88.6 86.5 87.6 0.8
5 152,411 123,738 128,735 121,284 81.2 84.5 79.6 81.8 2.0
6 159,727 131,957 140,593 134,098 82.6 88.0 84.0 84.9 2.1
7 157,171 140,362 152,411 146,913 89.3 97.0 93.5 93.2 2.7
8 168,173 148,766 157,665 159,727 88.5 93.8 95.0 92.4 2.3
9 156,905 157,171 162,919 158,786 100.2 103.8 101.2 101.7 1.5
10 168,830 161,057 168,173 157,846 95.4 99.6 93.5 96.2 2.1
* Percentage increases in market value for years 1 through 10, respectively, are: 7.3, 9.0, 3.7, 8.8, 15.5, 4.8, -1.6, 7.0, -6.7, and 7.6. 

Source: Author's calculations.

 

Table T5 Triennial Assessment without Phase-in - Hypothetical Example of Effects on Assessment Level and Uniformity Using Homes with Base Period Market Value of $100,000 and Experiencing Annual Value Changes at the Actual Annual Percentages Rates for Aggregate Residential Assessed Values for 1986-96, by Triennial Assessment Group

Year Market Value* 
(MV)
Assessed Value (AV) Ratio: AV/MV (Percent) Coefficient 
of 
Dispersion
Group 1 Group 2 Group 3 Group 1 Group 2 Group 3 Mean
Base $ 100,000 $ 100,000 $ 100000 $ 100,000 100.0 100.0 100.0 100.0 0.0
1 107,300 107,300 100,000 100,000 100.0 93.2 93.2 95.5 2.4
2 116,957 107,300 116,957 100,000 91.7 100.0 85.5 92.4 5.3
3 121,284 107,300 116,957 121,284 88.5 96.4 100.0 95.0 4.0
4 131,957 131,957 116,957 121,284 100.0 88.6 91.9 93.5 4.1
5 152,411 131,957 152,411 121,284 86.6 100.0 79.6 88.7 7.9
6 159,727 131,957 152,411 159,726 82.6 95.4 100.0 92.7 6.1
7 157,171 157,171 152,411 159,726 100.0 97.0 101.6 99.5 1.5
8 168,173 157,171 168,173 159,726 93.5 100.0 95.0 96.1 2.3
9 156,905 157,171 168,173 156,905 100.2 107.2 100.0 102.5 2.5
10 168,830 168,830 168,173 156,905 100.0 99.6 92.9 97.5 2.4
* Percentage increases in market value for years 1 through 10, respectively, are: 7.3, 9.0, 3.7, 8.8, 15.5, 4.8, -1.6, 7.0, -6.7, and 7.6. 

Source: Author's calculations.


Ending the Triennial System

Suppose the triennial system with phase-in is implemented, and the District later decides to abandon it in favor of annual adjustment of values aided by CAMA, coupled with less frequent property inspections. What would occur in the transition? In exploring this question, table T4 is the starting point. That table uses the actual annual percentage changes in market value over a 10-year period and incorporates the District's phase-in feature.

One concern raised by this transition might be the different experiences of the various triennial groups. Suppose the change came three years after implementation - i.e., after completion of the first triennial cycle for group 1. Several differences across groups would result (Table T6).

  • Different magnitudes of change in assessed value between years 3 and 4: The one-year change in assessed value ranges from a low of $20,652 (18.6 percent) for group 2 to a high of $24, 862 (23.2 percent) for group 3;
  • Different average levels of tax base over the life of the triennial system: The average assessed value over years 1-3 ranges from $102,365 for group 3 to $105,622 for group 2; and
  • Different average ratios of assessed value to market value (AV/MV) over the life of the triennial system: Average AV/MV for years 1-3 ranges from 88.9 percent for group 3 to 91.7 percent for group 2.

Because of the lag in posting value changes under the phase-in feature of the District's triennial system, a simultaneous move of all groups from triennial assessment to current market value assessment results in one-year increases that are a multiple of the one-year change in market value. Thus, while market value increases only 8.8 percent in year 4, assessed values for the three groups rise at two to three times that rate. The large one-year increase itself may pose some difficulty. The variation in the increases across groups may be an additional concern. This is made clearer by reference to the figures for average assessed value over the life of the triennial system (three years in this example). The three-year average assessment level for group 2 is $3,257 higher than for group 3. Thus, the average annual tax bill will have been $31 higher ($3,257 * .0096), using the current class 1 tax rate of 0.96 percent. (The same effects would occur for other classes of property, of course, and their higher tax rates under the current classification would result in larger dollar differences in average tax bills across properties having the same market value.) The average difference over the three years is only 2.8 percent in this example. Because the Board of Real Property Assessments and Appeals will not hear a complaint alleging a valuation error of under 5.0 percent, this may not be of great concern. The level of difference - i.e., the degree of inequity - will differ depending upon the pattern of annual market value changes and the point in the triennial cycles at which the change to current market value is made.

A second example is given in table T6. It is on the same basis as the first - i.e., the actual market value percentage changes for the period 1986-96 and triennial assessment with phase-in of value changes, shown in table T4. The difference is that the change in valuation systems is assumed to take place after seven years under the triennial approach. As before, coming off the triennial system with phase-in of value changes results in one-year changes in assessed value far greater than the one-year change in market value. Market value is up just 7.0 percent between year 7 and year 8, but assessed values changes are up anywhere from 10.3 percent (group 2) to 19.8 percent (group 1). Compared to the first example in table T6, however, the differences in the second example are greater - i.e., no two groups are close to the same percentage change in assessed value in the second example. The extremes of seven-year average assessment levels (AV/MV) differ by 3.4 percentage points (87.0 percent for group 3 and 90.4 percent for group 2), reflecting a $4,621 difference in average assessed value; this would result in an average annual tax bill difference of $44 at the current class 1 rate of 0.96 percent. The percentage difference between low and high average assessed values of the seven years is 3.9 percent.  

Table T6 Moving from Triennial Assessment System with Phase-in of Valuation Changes to an Assessment at Current Market Value, Selected Points in Triennial Cycle, by Triennial Group

Year Event or Item Market Value (MV) Assessed Value (AV)
Group1 Group 2 Group 3
End triennial after year 3, return to current market value standard in year 4
3 Last year of triennial 121,284 107,300 111,305 107,095
4 First year, current MV 131,957 131,957 131,957 131,957
  Assessed value change   24,657 20,652 24,862
  Percent change 8.8 23.0 18.6 23.2
  Average AV, years 1-3 115,180 104,867 105,622 102,365
  Average AV/MV, years 1-3   91.0 91.7 88.9
End triennial after year 7, return to current market value standard in year 8
7 Last year of triennial 157,171 140,362 152,411 146,913 
8 First year, current MV   168,173 168,173 168,173
  Assessed value change   27,811 15,762 21,260
  Percent change 7.0 19.8 10.3 14.5
  Average AV, years 1-7 135,258 118,025 122,275 117,654
  Average AV/MV, years 1-7   87.3 90.4 87.0
Source: Author's calculations based on table T4 and assumptions spelled out above.

Back to top of page


Send mail with questions or comments to webmaster@dcwatch.com
Web site copyright ©DCWatch (ISSN 1546-4296)