Cost-Benefit Analysis of Expanding Benefit Payment Control Activities
Ideally, states would prevent overpayment errors. In reality, however, UI payment administration staff are at the mercy of claimants' self-disclosures to be aware of most potential disqualifying eligibility issues. Claimants' reticence about such matters is behind the BAM conclusion that about three quarters of overpayments are "undetectable" to staff making payment determinations. States have found it is often more cost-effective to put their effort into detecting, establishing and recovering overpayments after the fact than into additional efforts to prevent them. In FY 2001, BAM estimates that about $2.3 billion was overpaid. In that same year, states reported that they established about $670 million through Benefit Payment Control (BPC) operations, and recorded about $360 million in recoveries. This exploratory analysis merges overpayment data from the ETA 227 report, BAM overpayment estimates, and BPC cost data from the new Resource Justification Model (RJM) to estimate the cost-effectiveness of existing BPC operations. It suggests a way of determining where such operations might most profitably be expanded, and indicates the likely costs and benefits of expanded efforts to increase the establishment and recovery of overpayments.
Basic State Data for FY 2001
Table 1 gives basic BPC data by state for FY 2001: Dollars established, dollars recovered, BPC staff years, and $ spent on BPC. It also shows dollars paid and the BAM estimate of dollars overpaid that corresponds most directly to BPC operations: fraud and nonfraud recoverable overpayments that are most detectable by BPC. This measure has been proposed for a GPRA accuracy measure and is thus referred to as the "GPRA Rate." These figures are for the year ending March 31, 2001. The six-month lag between the BAM estimate and the BPC establishment and recovery figures is used to reflect the typical lag between the time overpayments occur and BPC is able to detect them through the wage-record crossmatch, the most powerful tool for detecting recoverable overpayments. The table excludes PuertoRico (it did not report BPC data on the 227) and Virgin Islands (it has no BAM program).
These data indicate that in FY 2001, states recorded approximately $670 million in established overpayments and $360 million in recoveries. An approximate base for these activities is the $1.14 billion in most detectable, recoverable overpayments BAM estimated for the year ending six months earlier. As part of the RJM exercise, states reported devoting 1,446 staff years to BPC activities, at a cost of $72.9 million.
Table 1
| Payments and Overpayments 4/1/00-3/31/01 | BPC Activity and Cost Data, FY 2001 | |||||||
| $ Paid | GPRA Rate | GPRA OP$ | Established | Recovered | BPC SY | Cost/SY | BPC $ | |
|---|---|---|---|---|---|---|---|---|
| AK | $101,244,210 | 7.64% | $7,730,622 | $4,150,101 | $2,333,967 | 14.44 | $49,644 | $716,859 |
| AL | $211,973,506 | 4.13% | $8,754,366 | $7,104,893 | $4,202,761 | 23.22 | $40,334 | $936,555 |
| AR | $210,821,685 | 10.19% | $21,480,877 | $3,991,365 | $1,386,005 | 12.19 | $39,797 | $485,125 |
| AZ | $164,905,718 | 6.30% | $10,385,916 | $5,867,404 | $2,917,074 | 15.93 | $34,069 | $542,719 |
| CA | $2,489,220,165 | 3.67% | $91,247,595 | $119,473,837 | $57,805,515 | 238.32 | $59,971 | $14,292,289 |
| CO | $155,714,548 | 5.74% | $8,938,974 | $6,101,551 | $3,360,703 | 25.86 | $50,982 | $1,318,395 |
| CT | $346,333,861 | 2.56% | $8,879,629 | $6,283,364 | $4,694,541 | 17.99 | $90,989 | $1,636,892 |
| DC | $53,538,978 | 13.29% | $7,116,036 | $2,226,439 | $917,815 | 3.01 | $46,427 | $139,745 |
| DE | $72,153,915 | 7.49% | $5,407,650 | $1,473,077 | $1,053,656 | 8.25 | $39,365 | $324,761 |
| FL | $677,859,391 | 2.75% | $18,653,960 | $18,611,676 | $6,229,010 | 34.34 | $39,844 | $1,368,243 |
| GA | $333,568,914 | 2.15% | $7,182,661 | $6,539,497 | $3,486,061 | 37.61 | $47,377 | $1,781,849 |
| HI | $101,506,635 | 1.77% | $1,796,681 | $959,566 | $548,928 | 9.33 | $43,766 | $408,337 |
| IA | $236,283,733 | 5.47% | $12,914,609 | $8,506,763 | $5,447,070 | 15.39 | $49,181 | $756,896 |
| ID | $112,562,548 | 8.38% | $9,437,397 | $2,863,163 | $1,958,524 | 4.01 | $46,860 | $187,909 |
| IL | $1,336,133,878 | 8.49% | $113,488,565 | $59,415,800 | $32,462,841 | 68.89 | $61,787 | $4,256,506 |
| IN | $382,449,768 | 1.85% | $7,076,370 | $12,811,123 | $6,143,304 | 19.57 | $37,095 | $725,949 |
| KS | $172,476,048 | 3.12% | $5,388,732 | $3,993,502 | $2,684,681 | 11.97 | $40,714 | $487,347 |
| KY | $297,124,937 | 4.29% | $12,740,631 | $3,189,621 | $2,262,915 | 8.11 | $36,713 | $297,742 |
| LA | $196,592,529 | 6.58% | $12,940,863 | $8,553,857 | $3,576,274 | 14.58 | $38,441 | $560,470 |
| MA | $823,544,309 | 3.88% | $31,913,691 | $9,933,670 | $9,123,640 | 27.84 | $55,599 | $1,547,876 |
| MD | $266,086,241 | 9.96% | $26,510,276 | $12,038,221 | $7,656,768 | 28.55 | $49,237 | $1,405,716 |
| ME | $78,097,345 | 11.74% | $9,171,531 | $1,851,719 | $1,435,960 | 11.32 | $48,649 | $550,707 |
| MI | $1,044,098,893 | 4.66% | $48,682,977 | $20,495,046 | $9,748,029 | 7.00 | $59,111 | $413,777 |
| MN | $428,047,638 | 7.22% | $30,884,619 | $11,602,529 | $7,687,024 | 22.99 | $56,096 | $1,289,647 |
| MO | $341,654,145 | 6.18% | $21,129,936 | $12,895,668 | $6,545,274 | 19.74 | $47,893 | $945,408 |
| MS | $134,991,826 | 5.74% | $7,755,208 | $4,547,336 | $2,619,326 | 13.14 | $39,015 | $512,657 |
| MT | $60,344,864 | 6.56% | $3,958,058 | $566,766 | $475,375 | 7.73 | $40,521 | $313,227 |
| NC | $551,306,272 | 3.75% | $20,648,156 | $15,144,995 | $9,509,234 | 37.85 | $44,448 | $1,682,357 |
| ND | $33,519,579 | 1.88% | $629,988 | $826,359 | $568,887 | 8.08 | $40,052 | $323,620 |
| NE | $61,605,390 | 7.12% | $4,386,294 | $2,049,346 | $1,575,554 | 7.77 | $38,941 | $302,572 |
| NH | $23,836,503 | 2.73% | $650,943 | $658,844 | $451,312 | 7.77 | $42,414 | $329,557 |
| NJ | $1,122,197,979 | 5.70% | $63,949,795 | $35,172,697 | $24,261,407 | 68.90 | $45,974 | $3,167,609 |
| NM | $73,353,023 | 4.37% | $3,206,072 | $1,884,753 | $1,505,344 | 7.27 | $41,150 | $299,161 |
| NV | $209,378,825 | 7.32% | $15,318,517 | $4,442,207 | $2,056,597 | 15.92 | $50,467 | $803,435 |
| NY | $1,804,364,112 | 2.73% | $49,277,222 | $62,924,347 | $26,312,369 | 106.66 | $53,881 | $5,746,947 |
| OH | $797,850,046 | 6.09% | $48,550,284 | $22,621,681 | $12,826,367 | 60.67 | $54,304 | $3,294,624 |
| OK | $122,504,735 | 1.40% | $1,713,188 | $2,968,936 | $1,479,098 | 16.81 | $43,322 | $728,243 |
| OR | $373,778,926 | 3.96% | $14,802,273 | $11,319,057 | $6,248,794 | 34.89 | $51,885 | $1,810,268 |
| PA | $1,591,014,465 | 6.47% | $102,935,318 | $29,213,598 | $13,187,960 | 74.22 | $49,368 | $3,664,093 |
| RI | $131,671,770 | 3.42% | $4,507,032 | $3,587,186 | $1,782,786 | 17.12 | $56,172 | $961,665 |
| SC | $230,224,013 | 7.40% | $17,028,203 | $5,476,411 | $4,145,777 | 22.31 | $39,385 | $878,679 |
| SD | $17,251,400 | 2.47% | $425,605 | $427,651 | $336,520 | 8.17 | $33,346 | $272,437 |
| TN | $401,773,990 | 5.50% | $22,101,560 | $9,590,584 | $3,911,334 | 29.59 | $38,850 | $1,149,572 |
| TX | $955,034,435 | 8.13% | $77,676,518 | $37,054,731 | $22,123,443 | 54.68 | $44,232 | $2,418,606 |
| UT | $118,496,841 | 8.57% | $10,149,402 | $5,108,983 | $2,607,373 | 8.00 | $51,436 | $411,488 |
| VA | $196,382,359 | 9.03% | $17,726,807 | $5,688,920 | $3,189,708 | 23.22 | $43,562 | $1,011,510 |
| VT | $43,880,381 | 3.13% | $1,374,299 | $1,421,815 | $861,289 | 7.70 | $45,209 | $348,109 |
| WA | $857,540,196 | 10.11% | $86,732,387 | $36,609,996 | $18,761,330 | 49.62 | $51,639 | $2,562,327 |
| WI | $602,277,164 | 4.32% | $26,046,877 | $15,730,795 | $10,835,130 | 40.21 | $46,886 | $1,885,286 |
| WV | $121,268,620 | 1.84% | $2,236,849 | $1,678,680 | $1,109,166 | 14.14 | $37,950 | $536,613 |
| WY | $26,229,278 | 4.72% | $1,238,618 | $512,943 | $263,902 | 3.23 | $43,187 | $139,494 |
| US | $21,296,070,530 | 5.38% | $1,144,880,637 | $668,163,069 | $358,676,176 | 1446.12 | $50,433 | $72,931,875 |
Key Performance Ratios
Table 2 shows performance and analytical ratios that are key to developing a cost-benefit analysis. These ratios are:
Dollars Established per Staff Year (Est/SY) for FY 2001. This is one of the basic establishment productivity ratios. It ratio ranges from a low of $52,344 in South Dakota to a high of nearly $3 million in Michigan. The Michigan figure is definitely an outlier, being over three times the next highest ratio (Illinois.) UI actuarial staff managing indicate that there was a lack of detail in the Michigan RJM submission and the BPC staff years had to be estimated.
Dollars Established per BPC Dollar (Est$/BPC$). This is the other basic establishment productivity ratio. It is highly correlated with $/SY and so South Dakota again is lowest at about 1.6, and Michigan highest with an astronomical-and very suspect--49.5, due to the same estimation noted above. Fifteen states have ratios of 10 or more; fourteen are under 5.0.
Dollars Recovered per Staff Year (Rec/SY) for FY 2001. Recoveries are the "bottom line" to the agency; establishments are an intermediate product. This ratio averages $325,000 for the U.S. and ranges from $41,000 in South Dakota to nearly $490,000 in Idaho (if one ignores Michigan).
Dollars Recovered per BPC Dollar. This is the analogous productivity ratio to Est/SY for establishments. Every state recovers more than it spends on BPC activities, although for a few states (SD, HI, MT, NH) the ratio is very close to 1.0. Again, Michigan is the outlier with a ratio of 23.4, over twice the ratio calculated for the next highest state, Idaho, at 10.4.
Establishment Penetration for FY 00-01 (00-01 Est Pene). This is the ratio of actual established dollars to the BAM-estimated base. Because both numerator and denominator can fluctuate considerably (BAM due to sampling variability, establishments due to a variety of factors) a 2-year average ratio is used. Five states have ratios exceeding the theoretical maximum of 1.0--the highest being New York at nearly 1.5-and four others exceed 0.9. These unlikely ratios probably result from a combination of (1) Very active and successful BPC programs; (2) a BAM program in the state that underestimates recoverable overpayments (this probably reflects either BAM's known tendency to underestimate Benefit Year Earnings errors or perhaps deficient BAM operations generally in the state. BAM data for these states are under review.); and (3) Some states may establish overpayments that are in categories we have excluded from our BAM base.
Recovery Ratio for FY 00-01. This is essentially a 2-year average of the old BPC performance measure, except that fraud and nonfraud recoveries are added together here. The U.S. average is 0.53; state performance ranges from 0.32 in Arkansas to 0.82 in Maine.
Recovery Penetration for FY 00-01. This is the recovery analogue to the establishment penetration ratio, the ratio of dollars recovered over the BAM most detectable and recoverable base. Mathematically, it is the product of the Establishment Penetration ratio and the Recovery ratio. In fact, the 6-month lag between the occurrence and recovery of overpayments is probably far too short, but use of a 2-year average probably mitigates some of the timing problems. In contrast to the Establishment Penetration ratio, these values are all "well behaved:" they range from 0.06 in Arkansas to 0.71 in Indiana. The U.S. average is 0.32.
Table 2
| Est/SY | Est$/BPC$ | Rec/SY | Rec$/BPC$ | 00-01Est Pene | 00-01R/E | 00-01 RecPene | |
|---|---|---|---|---|---|---|---|
| AK | $287,403 | 5.79 | $161,632 | 3.26 | 0.55 | 0.54 | 0.29 |
| AL | $305,982 | 7.59 | $180,997 | 4.49 | 0.83 | 0.59 | 0.49 |
| AR | $327,429 | 8.23 | $113,700 | 2.86 | 0.20 | 0.32 | 0.06 |
| AZ | $368,324 | 10.81 | $183,118 | 5.37 | 0.72 | 0.51 | 0.37 |
| CA | $501,317 | 8.36 | $242,554 | 4.04 | 1.31 | 0.48 | 0.63 |
| CO | $235,946 | 4.63 | $129,958 | 2.55 | 0.66 | 0.59 | 0.39 |
| CT | $349,270 | 3.84 | $260,953 | 2.87 | 0.77 | 0.67 | 0.52 |
| DC | $739,681 | 15.93 | $304,922 | 6.57 | 0.28 | 0.41 | 0.11 |
| DE | $178,555 | 4.54 | $127,716 | 3.24 | 0.27 | 0.67 | 0.18 |
| FL | $541,982 | 13.60 | $181,392 | 4.55 | 0.77 | 0.36 | 0.27 |
| GA | $173,877 | 3.67 | $92,690 | 1.96 | 1.23 | 0.51 | 0.63 |
| HI | $102,847 | 2.35 | $58,835 | 1.34 | 0.40 | 0.54 | 0.22 |
| IA | $552,746 | 11.24 | $353,936 | 7.20 | 0.79 | 0.60 | 0.47 |
| ID | $714,006 | 15.24 | $488,410 | 10.42 | 0.29 | 0.71 | 0.21 |
| IL | $862,474 | 13.96 | $471,227 | 7.63 | 0.63 | 0.49 | 0.31 |
| IN | $654,631 | 17.65 | $313,914 | 8.46 | 1.34 | 0.53 | 0.71 |
| KS | $333,626 | 8.19 | $224,284 | 5.51 | 0.69 | 0.64 | 0.44 |
| KY | $393,295 | 10.71 | $279,028 | 7.60 | 0.31 | 0.66 | 0.20 |
| LA | $586,684 | 15.26 | $245,286 | 6.38 | 0.68 | 0.40 | 0.27 |
| MA | $356,813 | 6.42 | $327,717 | 5.89 | 0.32 | 0.79 | 0.25 |
| MD | $421,654 | 8.56 | $268,188 | 5.45 | 0.49 | 0.65 | 0.32 |
| ME | $163,579 | 3.36 | $126,852 | 2.61 | 0.24 | 0.82 | 0.20 |
| MI | $2,927,864 | 49.53 | $1,392,576 | 23.56 | 0.46 | 0.47 | 0.22 |
| MN | $504,677 | 9.00 | $334,364 | 5.96 | 0.38 | 0.70 | 0.27 |
| MO | $653,276 | 13.64 | $331,574 | 6.92 | 0.71 | 0.47 | 0.33 |
| MS | $346,068 | 8.87 | $199,340 | 5.11 | 0.55 | 0.61 | 0.34 |
| MT | $73,320 | 1.81 | $61,497 | 1.52 | 0.24 | 0.72 | 0.17 |
| NC | $400,132 | 9.00 | $251,235 | 5.65 | 0.67 | 0.63 | 0.43 |
| ND | $102,272 | 2.55 | $70,407 | 1.76 | 0.92 | 0.71 | 0.65 |
| NE | $263,751 | 6.77 | $202,774 | 5.21 | 0.57 | 0.73 | 0.42 |
| NH | $84,793 | 2.00 | $58,084 | 1.37 | 0.92 | 0.66 | 0.61 |
| NJ | $510,489 | 11.10 | $352,125 | 7.66 | 0.53 | 0.65 | 0.35 |
| NM | $259,251 | 6.30 | $207,062 | 5.03 | 0.54 | 0.79 | 0.43 |
| NV | $279,033 | 5.53 | $129,183 | 2.56 | 0.29 | 0.46 | 0.14 |
| NY | $589,953 | 10.95 | $246,694 | 4.58 | 1.48 | 0.42 | 0.63 |
| OH | $372,864 | 6.87 | $211,412 | 3.89 | 0.51 | 0.52 | 0.26 |
| OK | $176,617 | 4.08 | $87,989 | 2.03 | 1.44 | 0.50 | 0.71 |
| OR | $324,421 | 6.25 | $179,100 | 3.45 | 0.58 | 0.53 | 0.31 |
| PA | $393,608 | 7.97 | $177,687 | 3.60 | 0.35 | 0.47 | 0.16 |
| RI | $209,532 | 3.73 | $104,135 | 1.85 | 0.88 | 0.49 | 0.43 |
| SC | $245,469 | 6.23 | $185,826 | 4.72 | 0.38 | 0.68 | 0.26 |
| SD | $52,344 | 1.57 | $41,190 | 1.24 | 0.97 | 0.77 | 0.75 |
| TN | $324,116 | 8.34 | $132,184 | 3.40 | 0.36 | 0.48 | 0.17 |
| TX | $677,665 | 15.32 | $404,598 | 9.15 | 0.40 | 0.66 | 0.26 |
| UT | $638,623 | 12.42 | $137,369 | 3.15 | 0.62 | 0.43 | 0.27 |
| VA | $245,001 | 5.62 | $111,856 | 2.47 | 0.32 | 0.62 | 0.20 |
| VT | $184,651 | 4.08 | $378,100 | 7.32 | 0.96 | 0.60 | 0.58 |
| WA | $737,807 | 14.29 | $269,464 | 5.75 | 0.46 | 0.51 | 0.24 |
| WI | $391,216 | 8.34 | $78,442 | 2.07 | 0.63 | 0.72 | 0.46 |
| WV | $118,719 | 3.13 | $81,703 | 1.89 | 0.77 | 0.65 | 0.50 |
| WY | $158,806 | 3.68 | $244,085 | 4.88 | 0.43 | 0.54 | 0.23 |
| US | $454,697 | 9.09 | $325,922 | 6.34 | 0.61 | 0.53 | 0.32 |
Costs and Benefits of Additional BPC Effort
Overpayments recovered are the primary bottom line of BPC activity (a secondary effect is prevention of future overpayments through deterrence, but we cannot measure that here). The information above provides most of what is needed to make some simple estimates, by state, of the funds recovered, and the attendant costs, of increasing BPC activity.
Recovering overpayments is the third stage of a three-part activity: detecting potential overpayments; establishing certain of those detected as actual overpayments; and finally recovering the established amount. This analysis does not involve the detection stage and focuses on dollars established and recovered. Overpayments must be established before they can be recovered. In accordance with their own internal priorities, states allocate their staff and complementary computer resources between detection/establishment and recovery activities. As a beginning point, this analysis assumes that additional recoveries come by increasing the penetration of the universe of overpayments through additional detection and establishment, and that the proportion of establishments recovered remain unchanged, as shown in Table 2. If a state can increase the scale of its BPC operations to the limit of what can be established without declines in its establishment or recovery productivity, the unit cost of expanding recoveries remains the same as the average reported in Table 2. For example, for every staff dollar Alaska spends on BPC it establishes $5.79, and recovers $3.26; it recovers $.54 of every $1 it establishes. (The validity of these assumptions is tested later in the paper.)
The analysis proceeds as follows, and Table 3 gives the basic calculations and results.-
First, the states are arrayed by their 00-01 average establishment penetration rate. The 00-01 averages, also shown in Table 2, are used to give a better picture of average performance because the base is subject to sampling variation and the establishments and recoveries can vary from year to year for a variety of reasons
-
Second, because of concerns that not all states' BAM estimates reflect all a state's recoverable overpayments that BPC could detect and potentially establish, we assume that if a state already has a penetration rate of at least 80%, it has reached its virtual maximum. This eliminates the five states with establishment penetration rates of at least 100% (Georgia, California, Indiana, Oklahoma, and New York) as well as six others with penetration between 80% and 97%.
-
Third, we assumed that states can increase their establishment penetration by one half of the difference between our assumed maximum of 80% and their 00-01 average. Thus Arkansas, which averaged only 20% penetration during 00-01, is assumed to be able to raise its penetration by 30 points, to 50%.
-
The increased establishment penetration is multiplied by the 2001 base to give the increase in dollars established.
-
The state's dollars established in 2001 per BPC staff year divided into increased dollars established gives the increase in BPC staff needed.
-
Multiplying the increased dollars established by the 00-01 ratio of recoveries to establishments gives the increase in $ established.
| OP Establishment Data | BPC Information | Recoveries | ||||||||||
| State | GPRA OP $ | 00-01 Pene |
Pene Incr | New Pene | $ Increase | Est/SY | Incr SY | Cost/SY | $ Increase | 00-01 R/E |
$ Recovered | Ben/ Cost |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AR | $21,480,877 | 0 | 0 | 1 | $6,488,355 | $327,429 | 20 | $39,797 | $788,619 | 0 | $2,091,466 | 3 |
| MT | $3,958,058 | 0 | 0 | 1 | $1,112,582 | $73,320 | 15 | $40,521 | $614,876 | 1 | $802,859 | 1 |
| ME | $9,171,531 | 0 | 0 | 1 | $2,560,980 | $163,579 | 16 | $48,649 | $761,643 | 1 | $2,108,887 | 3 |
| DE | $5,407,650 | 0 | 0 | 1 | $1,424,857 | $178,555 | 8 | $39,365 | $314,130 | 1 | $949,291 | 3 |
| DC | $7,116,036 | 0 | 0 | 1 | $1,861,996 | $739,681 | 3 | $46,427 | $116,871 | 0 | $756,217 | 7 |
| ID | $9,437,397 | 0 | 0 | 1 | $2,402,456 | $714,006 | 3 | $46,860 | $157,673 | 1 | $1,705,692 | 11 |
| NV | $15,318,517 | 0 | 0 | 1 | $3,882,292 | $279,033 | 14 | $50,467 | $702,166 | 0 | $1,802,864 | 3 |
| KY | $12,740,631 | 0 | 0 | 1 | $3,136,589 | $393,295 | 8 | $36,713 | $292,792 | 1 | $2,077,256 | 7 |
| MA | $31,913,691 | 0 | 0 | 1 | $7,654,931 | $356,813 | 22 | $55,599 | $1,192,800 | 1 | $6,044,467 | 5 |
| VA | $17,726,807 | 0 | 0 | 1 | $4,241,820 | $245,001 | 17 | $43,562 | $754,210 | 1 | $2,636,444 | 4 |
| PA | $102,935,318 | 0 | 0 | 1 | $23,385,543 | $393,608 | 59 | $49,368 | $2,933,114 | 0 | $11,026,780 | 4 |
| TN | $22,101,560 | 0 | 0 | 1 | $4,907,848 | $324,116 | 15 | $38,850 | $588,277 | 0 | $2,345,216 | 4 |
| SC | $17,028,203 | 0 | 0 | 1 | $3,587,393 | $245,469 | 15 | $39,385 | $575,590 | 1 | $2,450,212 | 4 |
| MN | $30,884,619 | 0 | 0 | 1 | $6,497,257 | $504,677 | 13 | $56,096 | $722,185 | 1 | $4,560,266 | 6 |
| HI | $1,796,681 | 0 | 0 | 1 | $359,808 | $102,847 | 4 | $43,766 | $153,114 | 1 | $194,551 | 1 |
| TX | $77,676,518 | 0 | 0 | 1 | $15,363,257 | $677,665 | 23 | $44,232 | $1,002,778 | 1 | $10,064,369 | 10 |
| WY | $1,238,618 | 0 | 0 | 1 | $229,704 | $158,806 | 1 | $43,187 | $62,468 | 1 | $123,336 | 2 |
| MI | $48,682,977 | 0 | 0 | 1 | $8,235,483 | $2,927,864 | 3 | $59,111 | $166,267 | 0 | $3,849,334 | 23 |
| WA | $86,732,387 | 0 | 0 | 1 | $14,604,801 | $737,807 | 20 | $51,639 | $1,022,187 | 1 | $7,505,662 | 7 |
| MD | $26,510,276 | 0 | 0 | 1 | $4,077,682 | $421,654 | 10 | $49,237 | $476,155 | 1 | $2,667,741 | 6 |
| OH | $48,550,284 | 1 | 0 | 1 | $7,130,035 | $372,864 | 19 | $54,304 | $1,038,419 | 1 | $3,720,978 | 4 |
| NJ | $63,949,795 | 1 | 0 | 1 | $8,479,056 | $510,489 | 17 | $45,974 | $763,613 | 1 | $5,506,975 | 7 |
| NM | $3,206,072 | 1 | 0 | 1 | $419,257 | $259,251 | 2 | $41,150 | $66,547 | 1 | $332,111 | 5 |
| AK | $7,730,622 | 1 | 0 | 1 | $978,677 | $287,403 | 3 | $49,644 | $169,050 | 1 | $525,548 | 3 |
| MS | $7,755,208 | 1 | 0 | 1 | $966,455 | $346,068 | 3 | $39,015 | $108,956 | 1 | $590,843 | 5 |
| NE | $4,386,294 | 1 | 0 | 1 | $500,901 | $263,751 | 2 | $38,941 | $73,955 | 1 | $364,575 | 5 |
| OR | $14,802,273 | 1 | 0 | 1 | $1,593,379 | $324,421 | 5 | $51,885 | $254,831 | 1 | $848,611 | 3 |
| UT | $10,149,402 | 1 | 0 | 1 | $913,265 | $638,623 | 1 | $51,436 | $73,556 | 0 | $395,290 | 5 |
| IL | $113,488,565 | 1 | 0 | 1 | $9,906,639 | $862,474 | 12 | $61,787 | $709,705 | 0 | $4,887,918 | 7 |
| WI | $26,046,877 | 1 | 0 | 1 | $2,158,828 | $391,216 | 6 | $46,886 | $258,729 | 1 | $1,560,977 | 6 |
| CO | $8,938,974 | 1 | 0 | 1 | $603,535 | $235,946 | 3 | $50,982 | $130,409 | 1 | $356,152 | 3 |
| NC | $20,648,156 | 1 | 0 | 1 | $1,313,524 | $400,132 | 3 | $44,448 | $145,911 | 1 | $833,888 | 6 |
| LA | $12,940,863 | 1 | 0 | 1 | $772,847 | $586,684 | 1 | $38,441 | $50,639 | 0 | $306,062 | 6 |
| KS | $5,388,732 | 1 | 0 | 1 | $308,132 | $333,626 | 1 | $40,714 | $37,603 | 1 | $196,553 | 5 |
| MO | $21,129,936 | 1 | 0 | 1 | $918,219 | $653,276 | 1 | $47,893 | $67,317 | 0 | $431,308 | 6 |
| AZ | $10,385,916 | 1 | 0 | 1 | $414,674 | $368,324 | 1 | $34,069 | $38,356 | 1 | $211,410 | 6 |
| WV | $2,236,849 | 1 | 0 | 1 | $38,299 | $118,719 | 0 | $37,950 | $12,243 | 1 | $24,824 | 2 |
| FL | $18,653,960 | 1 | 0 | 1 | $310,981 | $541,982 | 1 | $39,844 | $22,862 | 0 | $111,078 | 5 |
| CT | $8,879,629 | 1 | 0 | 1 | $116,507 | $349,270 | 0 | $90,989 | $30,352 | 1 | $78,485 | 3 |
| IA | $12,914,609 | 1 | 0 | 1 | $84,547 | $552,746 | 0 | $49,181 | $7,523 | 1 | $50,537 | 7 |
| AL | $8,754,366 | 1 | 0 | $305,982 | $40,334 | 1 | ||||||
| RI | $4,507,032 | 1 | 0 | $209,532 | $56,172 | 0 | ||||||
| ND | $629,988 | 1 | 0 | $102,272 | $40,052 | 1 | ||||||
| NH | $650,943 | 1 | 0 | $84,793 | $42,414 | 1 | ||||||
| VT | $1,374,299 | 1 | 0 | $184,651 | $45,209 | 1 | ||||||
| SD | $425,605 | 1 | 0 | $52,344 | $33,346 | 1 | ||||||
| GA | $7,182,661 | 1 | 0 | $173,877 | $47,377 | 1 | ||||||
| CA | $91,247,595 | 1 | 0 | $501,317 | $59,971 | 0 | ||||||
| IN | $7,076,370 | 1 | 0 | $654,631 | $37,095 | 1 | ||||||
| OK | $1,713,188 | 1 | 0 | $176,617 | $43,322 | 1 | ||||||
| NY | $49,277,222 | 1 | 0 | $589,953 | $53,881 | 0 | ||||||
| US | $1,144,880,637 | 1 | $153,943,392 | $454,697 | 368 | $50,433 | $17,458,489 | 1 | $87,097,032 | 5 | ||
-
The net result, for the 40 states, implies that an increase of about 368 BPC staff, at a cost of $17.5 million, could result in an increase in $154 million in establishments and $87 million in recoveries. The average would be about $5 recovered for every $1 spent on BPC. As is shown below, this should be considered an upper limit for this ratio.
Testing the Realism of the Underlying Assumptions
The analysis above makes the critical assumption that BPC productivity will remain unchanged, despite the increase in the scale of its activity. In particular,
Staff will continue to establish and recover the same level of $ per staff year of activity as in FY 2001, and
Staff will be able to recover the same proportion of $ established as they did on average for FY 2000-2001.
Economic theory, however, assumes that after some point, productivity is subject to diminishing returns. In this case, the theory would predict that states would experience some declines in establishments per staff year, or in the ratio of recoveries per staff year, or in the ratio of recoveries to establishments as they increased the scale of their BPC activity. To test the validity of these assumptions directly, we would need data on individual state productivity ratios over time for different levels of BPC activity. We lack information on BPC staff usage over time. However, the data in Table 2 show that states vary widely in the proportion of potential overpayments they establish, from a high of 1.48 in New York to a low of 0.20 in Arkansas. This permits some indirect tests, by examining whether productivity of states with high establishment penetration differs systematically from that of states with low establishment penetration.
Establishment Productivity. We compared the average Dollars Established/BPC Staff year for the 11 lowest penetration states with that of the highest penetration states (i.e., the 11 at 80% or above. The $ established was adjusted for the index of Average Weekly Benefit Amounts (AWBA) to ensure that the amounts were not biased by differences in average payment levels. The highest penetration states' average was about 86% of that in the lowest. Several regressions attempting to explain $ Established or $ Established/SY were also run. All regressions included as an explanatory variable the 00-01 establishment penetration rate. In most of them, the sign on the penetration rate variable was negative, although in only one case was it significant. However, on balance these results suggest that expanding the scale of states' BPC activity will result in somewhat lower productivity at the margin.
Recovery Productivity. In the same way as with establishment productivity, we compared $ Recovered/BPC staff year for the high 11 and low 11 establishment penetration states. The results were similar to establishment productivity; the high penetration states' productivity was about 11 percent lower than in the low-penetration states. Regression results were similar also. The signs were usually negative, but only one equation had a significant (and negative) coefficient. These results also suggest that productivity will fall for the extra activity.
Recovery/Establishment Ratio. The high-low comparison and regression results both suggested a slightly lower ratio for the high-penetration states. However, this ratio probably reflects administrative decisions to emphasize either establishment or recovery activity through the allocation of BPC staff, and so it is hard to say much about movements in this ratio.
On balance, it seems that the cost-benefit ratios, and dollars established and recovered per staff year through expanding BPC activity, are somewhat overstated, probably by 5 to 15 percent. Thus the cost-benefit ratios of about 5-to-1 (or 4.8-to-1 if Michigan is not included) should be lowered to a range of 4.0-to-1 or 4.5-to-1.
What Should States Do to Improve Establishments and Recoveries?
The regressions intended primarily to assess whether increased penetration of the overpayment universe would reduce BPC productivity included two variables intended to indicate whether active or controllable BPC measures produced yielded higher ratios of dollars established to potential recoverable overpayments, or of dollars recovered to dollars established.
Regressions designed to explain Dollars Established and Establishment productivity included as an explanatory variable the % of dollars established in 2001 by controllable methods, such as Benefit-Wage and other crossmatches. The coefficients were usually negative in sign and very small. This suggests that active measures do not systematically produce better results.
Similarly, regressions designed to explain Dollars Recovered and recovery productivity included as an explanatory variable the % of dollars recovered by cash, versus offset. As with the "controllable" percentage in the establishment regressions, the coefficients were usually negative in sign and very small. This suggests that active measures do not systematically produce better recovery results.
There is also a possibility that not all states report the breakouts of these establishment and recovery data accurately. It may be worth repeating this analysis after all states have implemented Data Validation and have adjusted their reporting accordingly.
These regressions give no guidance for states to do other than simply increase the effort they put into establishment and collections, using the same basic procedures they have been following. The regressions do indicate a significant relationship between the extent of BPC effort and both establishments and recoveries, and when the ratio of recoveries to establishments is included in equations explaining the level of dollars established, it has a significant negative sign. This suggests that states can only put more effort into recoveries at the expense of establishments.
Conclusions
The analysis above indicates that increasing BPC activity has a high payoff-about $5 recovered for every $1 invested in BPC without adjustments, and at least $4 for every $1 invested after allowing for some diminishing returns as higher penetration levels are attained. The most cost-effective approach would seem to be to concentrate additional resources in states that now have low establishment penetration. The analysis suggests that a targeted increase of $17 million in 40 states could result in roughly $150 million in additional establishments and $87 million in recoveries. Available statistical data do not give any specific guidance about what works to increase establishments (e.g., use more active methods) or increase recoveries (e.g., concentrate on cash recoveries).
