Like any complex bureaucratic process, a public procurement system can be “gamed,” its rules manipulated to defeat the system’s purpose. Procurement systems are particularly susceptible to gaming for they are designed to advance two objectives in conflict. One is to allow governments to buy what they need when they need it quickly and easily. The second is to prevent fraud and corruption from infecting the system by imposing elaborate safeguards at every step in the purchasing process – at the cost of making it slow, cumbersome, and costly. The pressures to privilege the first at the expense of the second are many: the agency needs a replacement part immediately; every day the road is left unrepaired traffic snarls and citizens’ patience tested; overworked staff don’t have time to conduct a full-blown procurement. The result is that procurement officers are always on the lookout for ways to bypass, or “game,” the rules that slow the process down.
One way is to attach an unrealistically low estimate on what the item to be procured will cost. If the estimated price is below a certain amount, procurement officers can avoid conducting a full-fledged, open tender. Below the threshold, in many systems $1 million, procurement officers need not prepare a lengthy, formal tender document, advertise it widely for a several week period, constitute a technical committee to evaluate the bids, and follow the many other rules for open, competitive procurements. They can instead use streamlined procedures — variously termed “shopping” or a “request for” or “invitation to submit” quotes—which allow them to call a few suppliers for a price quotation and take the lowest one offered.
No one with experience in public procurement doubts that threshold gaming sometimes occurs. The questions are how often and why. Do procurement staff regularly underestimate the contract price to push it below the threshold and avoid the panoply of procurement rules that would otherwise have to be applied? Do staff do so to secure desperately needed items faster and cheaper? For other legitimate ends? Or to further corrupt deals?
New research now settles these questions – at least for the Czech Republic. Moreover, in answering them the researchers use techniques that others can employ to analyze the same questions in their countries.
The scholars in question are Ján Palguta and Filip Pertold of the Center for Economic Research and Graduate Education-Economic Institute in Prague. What they do is take advantage of a 2006 change in Czech procurement law to analyze the threshold gaming question. Upon joining the European Union the Czech Republic had to conform its national procurement law with the EU’s which required, among other changes, that a threshold be set below which goods, services, and simple, low-cost civil works could be bought by soliciting price quotes from five different suppliers. For 2010, the last year studied, the threshold was roughly $1 million for civil works and at the national level $160,000 for goods and services.
Palguta and Pertold examined the distribution of contract prices for goods, works and services around these levels for 2005 and compared that distribution to those in the years 2006 through 2010. What they found in these later years was a significant increase in the number of contracts falling just below the threshold amounts: for civil works (small construction) contracts the increase was 156 percent, for goods 113 percent, and for services 182 percent. A variety of tests confirmed that this bunching right under the threshold occurred; the tests also revealed that it was far more pronounced for construction and service contracts than those for goods. They ascribe this difference to the greater ease with which estimates for construction and services can be manipulated. If a procurement calls for the purchase of 100 laptop computers with certain specifications, an unrealistically low estimate is easy to spot. By contrast, it is much harder to observe whether the estimate to repair a road or provide consulting services has been artificially depressed.
As important as confirming the widespread suspicion that estimated contract prices are frequently gamed so that simplified procedures can be used was their finding about the purpose of gaming. Procurement officers sometimes claim in off-the-record discussions that they game the rules to cut costs and speed the acquisition of needed items. That whatever the ethics of gaming, the motive is benign.
The Czech government records the name of the firm winning each procurement, and Palguta and Pertold linked that information to data in the corporate registry showing whether the owners of the firm were publicly disclosed or anonymous (possible until recently because Czech law permitted the use of allowed “bearer shares,” stock certificates whose owners were not listed in the corporate registry.) As the authors argue, the award of a public contract to an anonymously owned firm increases the risk of corruption: the procurement officer may have a secret interest in the firm or it may be owned by politically connected individuals. And whatever the case, handing out public contracts to anonymously owned firms reduces the transparency of the process and casts a pall over its fairness.
Their findings strongly suggest that in the Czech Republic it is highly unlikely that procurement thresholds are gamed for benign reasons. After the introduction of simplified procedures, the chances that an anonymously owned firm would win a construction contract just below the $1 million threshold tripled. There was also an increase, although smaller, in the probability that an anonymous firm would win a procurement for a service contract just below the $160,000 threshold for that type of contract.
Further evidence that the threshold was gamed for corrupt ends comes from their comparison of the estimated contract price to the price the government ultimately paid: “Procurements from anonymously owned firms have higher contract prices compared to procurements with the same anticipated value below the thresholds but awarded to firms with traceable owners” (311). For works contracts the increase was close to nine percent; for services contracts it was almost six percent.
Palguta and Pertold provide important evidence on how procurement rules can be abused to foster corruption. One thing it shows is the importance of disclosing the beneficial owners of the companies winning public contracts. A second is the importance of gathering data across the procurement process so analysts can replicate their work in other countries. Finally, where as they say, a country has weak public institutions, policymakers should rethink procedures that allow procurement officers to sidestep the safeguards that come when a full-blown, open procurement is conducted.