Lowest Price Technically Acceptable (LPTA) may be the most feared term in the government today. The process, as outlined in the Federal Acquisition Regulations (FAR), was designed around a simple premise. That is, when the market offers functionally equivalent goods, the government should base its purchasing decisions solely on the lowest price.
The genesis of this contracting vehicle was noble. When goods are commoditized, trying to find the “best value” often results in diminishing returns. Procurement officers shouldn’t have to try and distinguish the value of one pencil over another. The government should just pick the cheapest standard pencil and buy it.
Yet in the current environment of sequestration and austerity, LPTA is metastasizing far beyond the FAR’s original intent. Procurement officers now commonly use LPTA to acquire goods that traditionally demanded a “best value” approach. The government has simply stopped looking at the tradeoffs in quality that impact mission effectiveness. As sequestration becomes the new normal, LPTA too is becoming an ingrained habit rather than a rare exception.
Data on LPTA procurements appear to confirm this trend toward overuse. A 2013 GAO analysis of DOD spending noted that in 2010, DOD used LPTA in 26% of all procurements over $25 million. By 2013, that number was up to 36%. For procurements under $25 million, DOD used LPTA 45% of the time.
LPTA and Biometrics
The biometrics industry is far from immune from the LPTA trend. LPTA procurements are now common for biometric equipment, and show no sign of retreat. Government agencies assume that biometric hardware has become so commoditized that there is no longer any purpose in distinguishing between the value of different solutions and the effect of quality on mission effectiveness.
This is demonstrably false.
The fact is that quality matters a great deal in biometrics. More to the point, lack of quality in biometric equipment has a tangible, measurable cost which makes LPTA almost entirely inappropriate as a procurement vehicle.
Why Quality Matters
US government officials in the field regularly collect fingerprints for a wide variety of immigration, law enforcement and military purposes.
The quality of these fingerprint images is measured through a score developed by the National Institute of Standards and Technology (NIST). This quality differentiation alone should exempt fingerprint devices from all LPTA procurements as outlined in the FAR.
Unfortunately, the certification process for devices is disconnected from the image quality scores which those devices produce. Appendix F of the FBI’s EBTS 10.0 specification is a technical baseline for capture capabilities, not a measure of device quality. As a result, some Appendix F-certified devices produce high quality images which allow for highly accurate matching, while others produce low quality images which result in false negatives or mis-identifications.
Is the difference in image quality significant enough to matter? The data clearly demonstrate that it is.
The National Institute of Justice recently published a study in which it compared the matching performance of different biometric readers against a common set of exemplars. The Crossmatch® Guardian® outperformed all of the other devices by a significant margin, matching against 100% of the sample data set. For comparison purposes, a leading, competitive fingerprint reader matched against the sample data set 94.2% of the time.
That doesn’t seem like a big difference, but consider the volume of transactions which come in from the field to the US government’s back-end databases – millions per day. That 5.8% starts to add up real fast.
Even more than the quantity of errors that a lower quality biometric reader would produce, the qualitative aspect is even more significant. Every fingerprint that is mis-identified or results in a false reject is a missed opportunity. It means that someone with a criminal or terrorist past eludes the security dragnet. It means that someone is granted access to a secure facility, even when they have no access privileges. Biometrics are used in high security scenarios, meaning that each mismatch leads to unusually severe consequences.
Obscuring the True Cost of LPTA
There is a structural gap in US government biometrics programs which makes LPTA appear more cost effective than it really is. That gap is between the collectors of biometrics and the processors of biometrics. As a rule, operational elements which collect biometrics in the field for the US government are disconnected on a budgetary level from the operational elements which process those biometrics. CBP collects biometrics, but OBIM processes them – two separate budgets. SOCOM collects biometrics, BIMA processes them – two separate budgets.
This gap obscures the total cost of ownership for biometric equipment. When collecting agencies purchase low-quality biometric equipment, they effectively push the cost of matching to the processing agencies. The best example is “gray area” matches, in which the database algorithm cannot match a low-quality print without help from a fingerprint examiner. These manual processes are expensive and time-consuming, and are a direct result of low-quality scanners at the operational level.
Yet, since the budgets of collectors and processors are separate, word rarely gets back to the collecting agencies about the impact their cheap scanners have on the budget of the processors. Hence the prominence of LPTA procurements, in which one agency writes a check that another agency has to cash.
LPTA and Biometrics Don’t Mix
The market for fingerprint readers has certainly changed in the last several years. Prices continue to fall and competition among providers is stiff. Nonetheless, this broader market dynamic alone is not enough to justify LPTA procurements for biometric equipment. The difference in quality is both measurable and significant enough to justify “best value” procurements, even for smaller purchases. Appendix F certification alone is not enough to ensure that a device captures quality images.
LPTA has its place. For goods where the market is truly commoditized and the difference in quality is negligible or irrelevant, LPTA is probably the way to go. However, biometric equipment clearly does not fall into this category. The difference in quality is both measurable and significant in every use case, making LPTA inappropriate for any biometric procurement.