OVERVIEW. session. Section. government. for expansion, vendors for. between in-state. and Nic Rockler. Tom Kavet. From: Page 1

Transcription

1 Memor randum Kavet, Rockler & Associates, LLC 985 Grandview Road Williamstown, Vermont U.S.A. Telephone: Fax: Cellular: Website: m To: Steve Klein, Legislative Joint Fiscal Office From: Tom Kavet and Nic Rockler CC: Jason Aronowitz, Department of Finance and Management; Sara Teachout, JFO Date: February 28, 2014 Re: State Government Contracting Analysis Initial Review OVERVIEW This analysis was performed in accordance with Act 112 of the legislativee session. Section 1 of the Act requires the Secretary of Administration and the JFO legislative economist to design and implement a one-year pilot project to conduct an econometric analysis when evaluating government contracts for goods and servicess by the Department of Buildings and General Services, the Agency of Natural Resources, and the Department of Corrections. Section 2 requires the Secretary of Administration, working with the legislative economist, to submit a report to the House and Senate committees on Government Operations and on Natural Resources and Energy, the House committee on Corrections and Institutions, and the Senate committee on Institutions on the resultss of the econometric analysis pilot project and any further legislative or policy recommendations for expansion, adaptation, or elimination of econometric analysis in government contracting. The purpose of this analysiss is to quantify the differential benefit to the State of Vermont throughh the utilization of in-state vendors versus out-of-state by JFO economistss pursuant to this statute, including source data collection, contract database evaluation, econometric modeling, application issues and associated recommendations. SOURCE DATA COLLECTION AND EVALUATION The first step in assessing the value of modelling the differential economic impacts vendors for expenditures made by State government. This memo outlines the results of analysis performed between in-state and out-of-state contracting was to attempt to collect relevant data Page 1

2 on all State contracts. This database would allow assessment of the value of contracting that could be performed in-state and critical contract characteristics that affect net economic impacts, such as the vendor industry, size of contract and the local content of subcontractors that may be used to perform contracted work. The information requested for each contract is listed on the following page. Unfortunately, assembling this information was extremely difficult. There are various State entities letting the contracts with varying degrees of knowledge about the contracts and a general unwillingness to request additional information critical to this analysis from current vendors. Although we were provided with some information on 524 contracts, 90 of the contracts were delivered without reference to the questions posed, in a format other than that requested, and offering no detail as to the geographic home location of contractors or any information regarding the spatial distribution of contracting, subcontracting, payroll and materials/services inputs. Of the remaining 434 contracts reviewed, response rates on even the most basic information was subaltern. 37 of these failed to supply a description of goods or services purchased. Only 99 provided the NAICS code of the prime contractor, and of these, most were at a highly aggregated 2-digit level. The remaining 335 records offered no NAICS detail on contract activity, being either blank or not available, and some, though it is unclear exactly how many, were duplicates. Although we provided placeholders as likely NAICS codes in such cases, using information on the prime contractor obtained from either the contractors own websites or those providing corporate directories found on the web (with uncertain reliability), this was very time-consuming and of questionable accuracy. Among the useable contract records provided, there were 110 out-of-state contracts listed. Of these, 33 responded that no Vermont subcontracts or materials were used in performing the work, while fully 77 had no response. There were no contracts affirming the use any Vermont inputs of any kind on the part of out-of-state contractors. Review of these 110 out-of-state contracts also revealed that in 16 cases there were qualified Vermont firms suitable to bid the job, 3 cases which indicated no Vermont firms were qualified, 2 indicated that it was not clear, and for the remaining 89 contracts, there was either no response or not available was indicated. With respect to whether an out-of-state contractor had any Vermont payroll, only 2 provided the dollar value for FY13 and 108 were either blank or not available. Due to the poor response rate in compiling current contract information, it is not possible to estimate the percentage of current Vermont contracts that could be performed by in-state firms, the industries and activities associated with such work, or the economic impacts associated with performing such work in-state. Page 2

3 CONTRACT INFORMATION REQUESTED FOR ECONOMETRIC ANALYSIS Survey Source Contract # Vendor# Vendor Name Vendor State on Contract Contract Description Contract Maximum Amount Beginning Date Expiration Date Origin Code Buyer Status (Active or Open) VISION Account # Dept ID Fed Fund (Y/N) Original Contract Value Multi-Year Agreement (Y/N) Project Manager or Buyer Full Name If Out-of-State Vendor, are Vermont Subcontractors Used to Fulfill Any Portion of the Contract? Or, is Vermont the source of any portion of goods sold? If Yes, Percent of Total Expenditures to Fulfill Contract Paid to Vermont Subcontractors / for Vermontsourced goods? If Yes, Description of Vermont Subcontracting Work / Vermont Sourced Goods If Vermont Vendor, are Out-of-State Subcontractors / Goods Sourced Outside of Vermont Used to Fulfill Any Portion of the Contract? If Yes, Percent of Total Expenditures to Fulfill Contract Paid to Out-of-State Subcontractors / for Goods Sourced Outside of Vermont If Yes, Description of Out-of-State Subcontracting Work / Goods Sourced Outside of Vermont Could This Contract be Bid On or Fulfilled by a Vermont Entity? Have Vermont Entities Bid on this Contract in the Past 10 Years? If Yes, Have Vermont Entities Been Selected to Fulfill This Contract? If No, What are the Primary Reasons a Vermont Vendor was Not Selected? NAICS Code: See Sheet 2 Vermont-based company? (Y/N) Describe company's presence in Vermont Number of employees residing in Vermont % of employees residing in Vermont Total Vermont Payroll for Most Recent Fiscal Year Percent of Total Payroll in Vermont Data reported by: Company or State Project Manager / Buyer? Page 3

4 ECONOMETRIC MODELING 5 HYPOTHETICAL CASES In the absence of actual contract data detail, we constructed 5 hypothetical contracts designed to illustrate potential differences in economic impacts by industry sector, contract size and, in one case, facility location, and analyzed net State economic impacts using the Vermont REMI model 1 and a KRA fiscal model based on REMI outputs. These hypotheticals provide a demonstration of how an econometric model could be used to quantify an in-state vendor contract price differential (expressed as a percentage of an out-of-state bid) that would yield an equivalent economic impact, as measured by net State tax revenue. In these hypotheticals, we employed model specifications that represent a maximum impact differential by assuming that out-ofstate firms have no direct Vermont payroll, and that in-state firms behave according to REMI model defaults (based on historical experience) with respect to in-state expenditures. While these 5 hypothetical cases are not actual contracts, they are based on information that indicates they are in sectors in which substantial State expenditure occurs and in which in-state vendors may be available. They include a $4M expenditure in the banking sector, a $2.5M contract for nursing services, two variants on a $5.4M expenditure for ambulatory care (one in which the providing facility is located in-state and the other in which the providing facility is located out-of-state) and a $5.1M insurance carrier expenditure. The two model runs for each case (instate vs. out-of-state vendors) assume that the in-state vendor does not subcontract any work performed to out-of-state subcontractors and that the out-of-state vendor does not subcontract any work to in-state subcontractors. In actual contracts, detailed information on subcontractors, if any, is essential for accurate impact analysis, as is information on the exact types of services and materials that are to be provided within the terms of the contract. MODEL RESULTS Per the table on the following page, there are substantial economic benefits to the State when in-state vendors are used. For example, a $5.1M contract for insurance sector services could result in 36 additional Vermont jobs and more than $325,000 in additional State tax revenues when compared to the same contract let with an out-of- 1 This model is provided by Regional Economic Models, Inc. of Amherst, MA, and has been used by KRA, the Joint Fiscal Office and various State Agencies and Departments for more than 15 years. It is a dynamic regional input-output and behavioral econometric model that estimates direct, indirect and induced economic impacts. Fiscal impacts are estimated by KRA using REMI model output and Vermont revenue data maintained by the Joint Fiscal Office, Detailed REMI model background information is available at Page 4

5 HYPOTHETICAL IN-STATE VS. OUT-OF-STATE CONTRACT COMPARISONS MODEL INPUTS MODEL OUTPUT COMPARATIVE STATISTICS Additional Vermont Tax Contract Direct IN-STATE-VENDOR Additional Revenue if Value Employment OUT-OF-STATE VENDOR Assuming All Direct Employment Within Total GDP Total Vermont Tax In-State Sector Assumed Assuming No Direct Vermont Employment Vermont, Using REMI Default Values Difference Total GDP Employment Revenue as a Total Vermont Impacts Total Vermont Impacts In-State Difference Difference In-State Percentage of vs. as Percent In-State vs. Out-of-State Output Income GDP Output Income GDP Out-of-State of Contract vs. Out-of-State Contract Employment Employment Value Out-of-State Value Banking $ 4, , ,151 3,141 79% 25 $ % Nursing Services $ 2, , % 8 $ % Ambulatory Care a. Facility Located in VT $ 5, , , ,636 1,554 2,258 1,024 19% 16 $ % b. Facility Located Out of VT $ 5, ,636 1,554 2,258 2,050 38% 33 $ % Insurance Carriers $ 5, ,408 1,726 3,440 3,388 66% 36 $ % Page 5

6 state vendor. This tax revenue differential forms the basis for an equivalent contract price comparison and shows that a $5.4M bid from an in-state firm would represent about the same net cost to the State as a $5.1M bid from an out-of-state firm. In this case, the differential is 6.4%. Thus, an in-state bid that is as much as 6.4% above an out-of-state bid for these services would have an equivalent impact in terms of net costs to Vermont taxpayers. In the 5 hypothetical cases run, the equivalent contract price differentials range from 1.5% to 7.6%, depending upon the industry sector and exact contract content and specifications. Through this type of analysis, differentials for individual contracts and/or contract groupings could be produced, providing guidance on equivalent bids from in-state versus out-of-state vendors APPLICATION ISSUES Although this analysis shows promise as an unbiased way to quantify in-state contract preferences, legal issues have been raised in this process that require formal review before using this type of analysis in awarding State contracts. Contracts supported by federal funds may be particularly affected by legal limitations on preferential in-state awarding practices. For example, it was noted that in Maine an executive order issued by the governor (order ) that included scoring criteria evaluating bidders economic impact on the Maine economy and State revenues was found to be unacceptable to the Federal Highway Administration. Exactly why it was deemed unacceptable and whether there are changes in the scoring procedure that could make it acceptable remains to be investigated. Concerns have also been raised as to the possible effects this procedure could have on in-state bidding, if vendors knew the in-state preference factors being used to evaluate competitive bids. Some believe higher in-state bids could result from this. Others have also raised concerns about the information burden of providing the kind of detailed operational information necessary to assess economic impacts accurately and the possible discouraging effect this may have on potential bidders. It is our understanding that current RFPs now request most additional information required to run this kind of analysis. We have no information, however, on the completeness of the information we have received to date, none of which has been provided in spreadsheet form necessary for analysis and model input. Lastly, the costs of running both individual and group contract analyses are difficult to estimate until the quantity of contracts to be evaluated is determined and the level of detail available for model inputs is ascertained. Until the information requested as a part of this analysis is collected and assessed, it is impossible to do either. Page 6

7 RECOMMENDATIONS 1) Prior to any further work on this analytic approach, a review by the Office of the Attorney General should be conducted so as to insure the legality of any potential application of this analysis in the State contract bidding process. This is particularly important for contracts involving any federal funding. 2) If deemed legal, bidding information necessary to perform accurate economic impact modeling should be required by statute, so as to insure uniform, impartial, thorough and timely analysis. We do not believe this information collection process would represent an undue burden on the bidders or significantly affect the number of bidders available (see sample minimum RFP information form on following page). It would also make clear that the State is evaluating and favoring the use of both in-state vendors and the use of instate subcontractors with out-of-state vendors. Both will encourage the development of in-state contracting capabilities relevant to State needs. 3) After collecting relevant data on all State contracts, an analysis of the contracts by size, industry, function and possible in-state bidding potential should be conducted so as to be able to target analysis of relevant contracts and contract groups through this type of analysis. 4) At a minimum, large contracts and/or groups of similar smaller contracts with the potential to be served through in-state vendors should be analyzed with this methodology so as to provide objective quantification of equivalent instate contract price differentials. 5) Even if not formally used in the bidding process, individual State contracting agencies may wish to use this modelling approach to informally assess contract benefits from in-state contract awards. 6) A review of other states practices with regard to preferential in-state contract awarding processes should be undertaken. If there are other legal and objective ways of evaluating in-state benefits that are being used elsewhere, Vermont should also consider such approaches. Page 7

8 Sample Proposed RFP Information Form For contract amounts exceeding $50,000, bidders are requested to respond to the questions identified below. Act XXX requires measurement of the net fiscal impact to the state of certain identified purchases. In order to accomplish this goal, we are seeking data associated with contracts for goods and services with which to support this evaluation. The below questions will assist the state in estimating net state costs and benefits associated with identified contract bids. Please contact the contract administrator if you have any questions regarding this information. 1. Is your firm a Vermont-based company? Yes No 2. NAICS Code of primary contract activity (see attached list) 3. Describe Goods and services to be produced as a part of this contract: 4. Indicate the total number of employees expected to be working on this contract: (total employees) and the approximate percentage of these who reside in Vermont: (%). 5. Please indicate total expected payroll for this contract for all personnel ($ for wages, salaries and payroll taxes): (labor payroll $) and the percentage of this that will be earned by Vermont residents: (%). 6. Indicate the total value of all subcontracts that are included in your bid: ($) and the percentage of that value that will likely be paid to Vermont subcontractors: (%). 7. Please indicate the total value of supplies and services (not subcontracted) that are include in your bid amount: ($) and the likely percentage that will be paid to Vermont suppliers: (%). Page 8