In the January 31 Advocate, PHADA published a detailed analysis of HUD's Physical Inspection Scoring System showing that there were thousands of points which could be deducted, but only 41 were needed to fail. In addition, PHADA sent comments to REAC on fifty specific observable deficiencies which PHADA believes need to be changed. Prior to the publication of the analysis, PHADA had sent it to HUD for comments, but it received none.

REAC has recently posted a response on its web site to PHADA's analysis and comments in which it claims the analysis is based on a number of "erroneous assumptions" and that the PHADA analysis "may be misleading to members."

After reviewing REAC's response, PHADA does not believe it points out any "erroneous assumptions" in its analysis. Each point which HUD makes was fully discussed, explained and accounted for in PHADA's analysis and comments. The PHADA response to REAC which follows will review and disprove each one of REAC's points one by one. Therefore, PHADA contends that until proven differently, its analysis stands as complete and accurate.

PHADA is hopeful, however, that HUD is sincere in its response to PHADA's scoring analysis, when it says it welcomes a dialogue with its industry partners on the merits of the weights and values used in scoring the physical inspection system. One reason this dialogue is welcome is because in the past HUD has refused to discuss the weights and criticality levels.

PHADA Calculates Point Deductions Correctly
HUD's major contention -- that PHADA's point deductions are overstated -- comes in a table which compares PHADA's values, as found in its comments, versus HUD's values, which are presented as much lower. The fact is there is no real difference between PHADA's numbers and HUD's numbers.

The only reason they are different is that PHADA's numbers represent the amount of points deducted by a deficiency when scoring for each area, while HUD's represent the number of points deducted when scoring the overall development. PHADA made it clear to HUD in its comments that it was deducting points by area, "For each deficiency, PHADA will show what percentage of points its deduction is of the number of point deductions which would cause the authority to fail in that area. (emphasis added)". PHADA also made it clear in its comments that in its examples the area included only one building and only one unit.

Deducting points by area (or sub-area) is exactly the methodology used in HUD's published description of May 13, 1999, "Public Housing Assessment System Physical Condition Scoring Process". HUD states in Section 10, page 26168 of the Federal Register, that when one calculates the score for the Site area (or the other areas if they only include one building or one unit) one must "calculate the deduction for an observed deficiency by multiplying the relevant item weight by the criticality value and by the severity value".

PHADA's numbers are calculated exactly as described in HUD's Scoring Process published on May 13, 1999.

It follows up in Section 14, page 26168, using the example of a damaged roof vent, rated as minor, stating, "the amount of points deducted is the item weight (16.7) times the criticality value (3.00) times the severity value (0.25) which equals 12.5 points." It does not say to multiply by the percent value of the area, as REAC says to do in its response when it writes, "to approximate the actual points deducted for a given deficiency, that product (of the weight times the criticality times the severity) must also be ... multiplied by the area weight." PHADA's numbers are calculated exactly as described in HUD's Scoring Process.

Interpreting HUD's Table
HUD's table has a column entitled "HUD's Actual Points Deducted" which lists the points deducted for the overall development, while, as mentioned, the column entitled PHADA's Points Deducted lists the points deducted per area (assuming each area has one building or one unit). Thus, by comparing apples to oranges and not explaining what PHADA's points represent, HUD makes it appear as if PHADA has assigned too much value to its point deductions.

Although HUD's headline states that the PHADA "analysis may be misleading to members," it is HUD which is being misleading. Although PHADA clearly stated that its point deductions were for the area, HUD has tried to make its point deductions appear smaller and thus less punitive by multiplying them by the area's percentage value of the overall value of the development.

REAC's table merely takes the scoring process one step further, which is fully explained in PHADA's analysis
PHADA's Example No.	Inspectable Area	PHADA's Points Deducted	Method to Arrive at HUD's Actual Points Deducted	HUD's Actual Points Deducted	Method Used	HUD's Actual Points Deducted
4	Site	7.03	(7.03) x (.15)	1.1		1.1
16	Building Exterior	9	(9) x (.15)	1.4	(9) x (.15) / (5 bldgs)	0.3
28	Building System	46.5	(46.5) x (.20)	9.3	(46.5) x (.20) / (5 bldgs)	1.9
41(1st)	Dwelling Units	50	(50) x (.35) / 10 units	1.75	(50) x (.35) / (25 units)	0.7

Thus in PHADA's example #4, which accounts for a loss of 7.03 points in the Site area, HUD has merely multiplied 7.03 by .15, the value of the Site area, to arrive at its deduction of 1.1. Consequently, they are the same number. REAC has just carried the scoring one step further and listed the number of points deducted for the development. The same method is used for each example with HUD also dividing by the number of buildings or units where appropriate.

In its January 31 analysis, PHADA clearly stated in more than one instance, that after the area score is calculated, a PHA has to multiply the area score by the percent value of the Site to arrive at its portion of the total score. "Since the Site is worth 15 points of the total 100 for the development, the actual number of points the authority would receive for the Site would be the (Site total) X .15", the PHADA analysis states.

So, although HUD and PHADA both multiply the site score by the .15 and end up with exactly the same number, HUD has tried to make it appear as if there are fewer points than PHADA's analysis demonstrates by multiplying the area point deduction by the .15 value before the area score is calculated and then comparing its number with PHADA's area point deduction.

HUD's May 13, 1999 published scoring process, however, uses the same method PHADA has used, calculating the area score first and then multiplying by the area weight. On page 26168 of the Federal Register, Section 11 describes how to calculate area scores. Section 12 then reads, "to calculate the overall property score, the normalized area weights are applied to the area scores." Thus HUD's own method, published in the Federal Register, describes the system as first calculating an area score and then once it is derived, calculating the overall score. Since this method is the one presented in the Federal Register, it was perfectly logical for PHADA to present comments to HUD using points deducted per area. (An area with one building or one unit is calculated the same way as a sub-area).

REAC also tries to make a point by saying that the points deducted will be affected by the number of buildings or units inspected. In its analysis, however, PHADA specifically states the very same thing, "In Dwelling Units, one divides the number of points per observable deficiency by the number of units inspected," and then it proceeds to do just that in its examples.

Secondly, in its comments to REAC, although PHADA did not divide by units, it clearly stated that it would not and why. PHADA wrote, "This percentage (of amount needed to fail the area) is derived by assuming that the authority inspected has only ... one building, one unit. This assumption is valid, because if the deficiency was found in ... every building and every unit, the percentage of the total needed to fail would be identical." Thus PHADA clearly stated its assumptions and its reasons.

PHADA might well add that HUD uses this exact same method in its paper, "Physical Inspection Scoring" published on the REAC web site. This paper has a chart showing point losses by area assuming only one building or unit in just the same manner as PHADA has done, and just as PHADA has done, HUD writes, "It needs to be recognized that area scores are weighted averages over all buildings or units inspected. Thus, the impacts in Table 1 are accurate only if all buildings or units have deficiencies at the given level." Since HUD itself has published a paper using this methodology, it should not criticize PHADA for using it as well.

This explanation demonstrates that HUD is incorrect when it writes, "The PHADA analysis miscalculates points deducted."

The Concept of Losing 41 Points is a Valid Approach
Making another criticism, HUD writes that "41 points is not a valid basis for approach used." It finds fault with PHADA's writing that "deducting 41 points gives an authority a score of 59 which is the equivalent of failing and would require referral to the Troubled Agency Recovery Center (TARC)." There can be no doubt that scoring 59 on the physical inspection causes an authority to fail. In the PHAS final rule, HUD writes, "In order to receive a passing score under the Physical Condition Indicator, the PHA must receive a score of at least 60 percent of the available points under this indicator." It continues, "A PHA that achieves less than 60 percent of the total points under only one of the following indicators, #1 (physical) ... shall be considered a substandard physical ... performer and referred to the TARC."

In another section, HUD states that PHADA believes a PHA can fail an inspectable area. HUD surely recognizes that PHADA was using this imagery to make the complex scoring system easier to understand. It is also true that if a PHA fails all five areas, it will score below 60 for that development. Knowing the impact a single deficiency has on an area or sub-area's score is extremely relevant to knowing whether the point loss is appropriate to the deficiency's importance.

HUD itself uses this approach in its paper, "Physical Inspection Scoring" when it identifies deficiencies which have a high impact on an area score. Nevertheless, PHADA made it clear that a PHA does not fail an area, for it wrote, "Since points are deducted from the number 100, it only takes 41 points deducted to get a 59 or the equivalent (emphasis added) of a failing grade in the Site area...."

Large Point Deductions Are a Serious Problem
REAC is correct that PHADA decries large point deductions for single deficiencies. REAC has recognized this phenomena as well in the paper it published entitled "Physical Inspection Scoring." It writes, "for Site, Building Exterior and Building Systems, two medium impact deficiencies ... would ... deduct all or most of the possible points." This admission on HUD's part clearly shows PHADA is correct in decrying large point deductions. Note that in this paper (www.hud.gov/reac/pdf/reapscr.pdf) REAC is calculating area scores and deductible points for an area just as PHADA does.

This paper also publishes a table showing that even deficiencies regarded as minor can be categorized as medium impact. Thus, HUD acknowledges a point PHADA has made. PHAs can lose a disproportionate number of points for a minor deficiency. A minor deficiency should not be worth so many points that a few minor deficiencies cause a PHA to lose most or all of the points in an area. HUD's own table proves the value of determining the impact single deficiencies have on an area score.

It is clear that the area and item weights and criticality and severity levels do not always reflect the importance of the deficiencies. PHADA has proven this fact over and over. Unfortunately, REAC does not comment in its response on the many reasons PHADA has presented why the weights and criticality levels need to be reviewed. These include the fact that all the observable deficiencies in an item have the exact same weight, and the fact that a deficiency has the same criticality value, whether it is minor, major or severe.

Adding points up vs. deducting from 100
HUD then writes that PHADA's analysis suggests that the scoring system should add points for all items which are not deficient (the credit system) rather than deducting points. PHADA has never proposed the credit system, although this idea is worth exploring. HUD admits that if points were added up for doing well, virtually all PHAs would have scores in the 90s on a percent basis.

There is an important implication of this admission. By admitting that virtually all PHAs would score in the 90s percent-wise, HUD acknowledges PHADA's main point -- there are thousands of potential point deductions and that a scoring system which deducts points from the number 100 and fails PHAs below 60 is not related to a percent, but rather is completely arbitrary. Basically the scoring system is one in which HUD has chosen the number 41, apparently at random, and said losing 41 points causes a PHA to fail, no matter how many thousands of points are possible or how the PHA loses the 41 points.

PHADA's position remains that the current scoring system is deeply flawed and needs to be revised.

PHAS more stringent than "decent, safe and sanitary"
HUD says its deficiencies are necessary to make sure units are decent, safe and sanitary, but PHADA has also pointed out in an article in the March 13, 2000 Advocate that many deficiencies in PHAS go far beyond decent, safe and sanitary, which is the standard Housing Quality Standards (HQS) achieve. PHAs can fail for deficiencies which pass HQS.

PHADA is disappointed that HUD's response does not include one comment on the fifty specific deficiencies it brought to HUD's attention as minor, insignificant or disproportionate. Without a discussion of these important points, PHADA is concerned that the dialogue may become a monologue.

Physical Inspection Scores for Housing Authorities
Finally, HUD presents figures to show that the PHAS system is not unduly stringent. The fact that 78 percent of PHAs are either high performers or standard shows how well maintained public housing is, not that the system is fair. PHADA, however, finds these numbers much more disturbing than HUD does. If 78 percent of PHAs are either standard or high performers, it means 22 percent, or 700, PHAs would have been troubled in at least one area. In addition, there are close to 500 additional PHAs which scored between 60 and 70 on the physical inspection. PHADA does not believe 700 PHAs, or twenty times the current number, should be branded troubled based on a system that is so flawed. HUD's statistics also do not reflect the number of units found in these 700 PHAs. According to HUD charts, almost 40 percent of the nation's units are troubled in one area.

[REAC] says that if points were added for doing well, rather than deducted for deficiencies, "virtually all scores would be in the 90's" (on a percent basis). That is PHADA's point exactly. If over 90 percent of a housing authority's properties are well maintained, or if over 95, 96, 97 or even 98 percent are in good condition, is it fair to call it "troubled"?

In conclusion, contrary to HUD's assertion, PHADA has shown that its analysis and comments are not misleading. PHADA's comments clearly stated they were deducting area points, as instructed by HUD in the May 13, 1999 Physical Inspection Scoring Process. REAC has not made a single point which differs from PHADA's analysis or comments.

Instead, REAC has indirectly confirmed PHADA's main points. REAC acknowledges that its scoring system is extremely demanding, to the point PHADA believes it is unreasonable. It says that if points were added for doing well, rather than deducted for deficiencies, "virtually all scores would be in the 90's" (on a percent basis). That is PHADA's point exactly. If over 90 percent of a housing authority's properties are well maintained, or if over 95, 96, 97 or even 98 percent are in good condition, is it fair to call it "troubled"? The troubled designation is even less legitimate, because there are numerous deficiencies whose point deductions are excessive (PHADA has submitted twenty pages of comments on these deficiencies) and because the weights and criticalities have yet to be reviewed with PHAs and the industry groups. A score based on flawed deductions is not valid.

PHADA recognizes that creating this inspection system has not been an easy task. PHADA hopes that a continued dialogue can improve it to the point that it can accomplish HUD's goals of identifying authorities that need improvement and directing resources to meet identified needs.

PHADA analysis and comments are at: www.phada.org/phas.html. REAC's response to PHADA is at: http://www.hud.gov/reac/products/phas/phas_phada.html.

PHADA Advocate (vol. 15, number 5, 3/27/00)

PHADA FRONT >  ADVOCATE