RANDY T SIMMONS PH.D
Utah State University
RYAN M. YONK PH.D
Utah State University
TYLER BROUGH PH.D
Utah State University
KEN SIM
Strata Policy
JACOB FISHBECK
Strata Policy
RENEWABLE PORTFOLIO STANDARDS:
MICHIGAN
FINAL REPORT | SEPTEMBER 2015
RENEWABLE PORTFOLIO STANDARDS:
MICHIGAN
PRIMARY INVESTIGATORS:
Randy T Simmons, PhD
Utah State University
Ryan M. Yonk, PhD
Utah State University
Tyler Brough, PhD
Utah State University
Ken Sim, MS
Strata Policy
Jacob Fishbeck
Strata Policy
Institute of Political Economy, Utah State University
|
1
TABLE OF CONTENTS
Table of Contents ............................................................................................................................ 2!
Executive Summary ........................................................................................................................ 3!
Background ..................................................................................................................................... 3!
Results ............................................................................................................................................. 8!
Empirical Analysis ......................................................................................................................8!
State Coincident Event Study .....................................................................................................8!
The Structural Panel VAR-X Model .........................................................................................12!
Conclusions from the Empirical Analysis ................................................................................13!
Institutional Analysis ................................................................................................................14!
Conclusions from the Institutional Analysis .............................................................................19!
Conclusion .................................................................................................................................... 19!
Explanation of Empirical Study Methodology ............................................................................. 21!
0.1. The Panel VAR-X Model ..................................................................................................21!
0.2. Model Estimation ...............................................................................................................23!
0.3. Dynamic Multiplier Analysis .............................................................................................26!
Graphical Analysis of the Dynamic Multiplier Analysis .............................................................. 28!
2
|
Renewable Portfolio Standards: Michigan
EXECUTIVE SUMMARY
The U.S. has no federal mandate for “renewable” power production. Instead, a majority of states,
including Michigan, have created their own state laws called Renewable Portfolio Standards
(RPS). These laws mandate the presence of certain renewable sources among the overall menu of
sources from which electricity companies produce power. This report analyzes how the changes
in electricity markets caused by RPS alter the functioning of a state’s economy and institutions,
with a specific focus on Michigan. Our report uses an empirical analysis and a survey of legal
rules. The following are our key findings:
•! Our empirical analysis discovered significant harmful effects on the economies of all
states with RPS. States that have adopted an RPS have seen a drop in industrial electricity
sales by almost 14 percent. Real personal income has fallen by almost four percent,
which for Michigan figures to a loss of $15.1 billion in 2013, or $3830 less per family.
Non-farm employment has declined by nearly three percent. Lastly, RPS is correlated
with an increase of 10 percent in a state’s unemployment rate, equaling a loss of 24,369
jobs in Michigan.
•! Our analysis of the legal rules surrounding RPS in Michigan suggest that the regulatory
climate is burdensome for both utilities and bureaucracies, making RPS an even worse
venture for taxpayers than the tax-based or empirical analyses suggest.
BACKGROUND
Despite the existence of an extensive manufacturing infrastructure prime for the production of
renewable electricity equipment, the adoption of utility-scale renewable electricity generation
has been relatively sluggish in Michigan. By 2007, only 605 kilowatts (kW) of solar photovoltaic
Institute of Political Economy, Utah State University
|
3
capacity had been installed in the state,1 and only 55.5 megawatts (MW) of capacity had been
installed for wind. 2 In April of 2007, a study conducted by students at the University of
Michigan’s Center of Sustainable Systems analyzed different approaches for reducing GHG
emissions, one of which being RPS, and examined their potential economic impacts. They found
that a 20-percent RPS would significantly reduce GHG emissions while benefiting the Michigan
economy. 3 This set the stage for legislators to begin crafting renewable electricity incentive
programs.
Michigan’s experience with government-assisted renewable electricity begins before the RPS,
though. Legislators attempted to enact a feed-in tariff system for renewable energy sources,
similar to those used in Germany and Ontario, when Representative Kathleen Law introduced
HB 5218 in September of 2007.4 That bill stalled in committee, but Rep. Jeff Mayes and Rep.
David Palsrok introduced two similar bills in December—HB 5548 and HB 5549 respectively—
reviving that cause.5,6 In June of 2008, however, legislators also reinvigorated debate on an RPS
bill, SB 213, and around that time legislative efforts on the feed-in tariff bills effectively stopped,
1
Michigan Department of Labor & Economic Growth, Energy Office. (n.d.). Solar power installed in Michigan.
Retrieved from http://www.michigan.gov/documents/CIS_EO_Solar_Chart_140010_7.pdf
2
Michigan Department of Labor & Economic Growth, Energy Office. (n.d.). Wind power installed in Michigan.
Retrieved from http://www.michigan.gov/documents/CIS_EO_Wind_Chart_140011_7.pdf
3
Edison, M. H., Elliot, K., Fischlowitz-Roberts, B., Permut, R. A., Popp, S. A., & Winkelman, A. G. (2007, April).
Michigan at a climate crossroads: Strategies for guiding the state in a carbon-constrained world. Retrieved the
University of Michigan Library, Deep Blue web site:
http://deepblue.lib.umich.edu/bitstream/handle/2027.42/50484/MCCP_Final_SNRE_Report.pdf?sequence=1&isAll
owed=y
4
Michigan Legislature. (n.d.). House Bill 5218. Retrieved from
http://www.legislature.mi.gov/(S(0p3x0yej0taoxk1ivc4zwitt))/mileg.aspx?page=GetObject&objectname=2007-HB5218
5
Michigan Legislature. (n.d.). House Bill 5548. Retrieved from
http://www.legislature.mi.gov/(S(klizxy35opxrpzkjzzxqpenv))/mileg.aspx?page=getobject&objectname=2007-HB5548
6
Michigan Legislature. (n.d.). House Bill 5549. Retrieved from
http://www.legislature.mi.gov/(S(3smjhi3ksl5gsds0vguyspro))/mileg.aspx?page=getobject&objectname=2007-HB5549
4
|
Renewable Portfolio Standards: Michigan
allowing them to fade from discourse. 7 On October 6th, 2008, Governor Jennifer Granholm
signed SB 213 into law.8 Formally titled the Clean, Renewable, and Efficient Energy Act, or
Public Act 295 for short, that law is responsible for helping the state transition to a clean energy
economy. It does this by establishing a Renewable Energy Standard (RES)—Michigan’s version
of an RPS—requiring all investor owned utilities, municipal utilities, electric cooperatives, and
alternative electric suppliers to generate at least 10 percent of their electricity from renewable
sources, or to purchase and retire the equivalent amount in tradable renewable energy certificates
(RECs), which represent electrical generation.9
In 2012, a coalition of labor and industry groups called Michigan Energy, Michigan Jobs
proposed a constitutional amendment to establish a new RES benchmark of 25 percent
renewable electricity by 2025,10 notwithstanding that as early as 2011, utilities such as Detroit
Public Lighting and the City of Sebewaing reported that they would not be able to achieve
compliance with the 2015 RES benchmark. 11 The proposal appeared on the November 2012
general election ballot as Proposal 3, but was defeated with over 60 percent of Michigan voters
in opposition,12 leaving RES requirements unchanged.
The RES also includes annual interim compliance requirements, which started in 2012.13 In its
2011 report to the Legislature on the implementation of the RES, the Michigan Public Service
7
Michigan Legislature. (n.d.). Senate Bill 213. Retrieved from
http://www.legislature.mi.gov/(S(zf1chjlujjoatqibxwb2xonu))/mileg.aspx?page=GetObject&objectname=2007-SB0213
8
Ibid.
9
United States Department of Energy. (2014, October 28). Database of state incentives for renewables and
efficiency. Retrieved from http://programs.dsireusa.org/system/program/detail/3094
10
Michigan Secretary of State. (n.d.). Full text of proposed constitutional amendment. Retrieved from
http://www.michigan.gov/documents/sos/Full_Text_-_MI_Energy_MI_Jobs_399441_7.pdf
11
Michigan Public Service Commission. (2011, February 15). Report on the implementation of the P.A. 295
Renewable Energy Standard and the cost-effectiveness of the energy standards. Retrieved from
http://www.michigan.gov/documents/mpsc/Report_on_Implementation_of_PA_295_RE_345746_7.pdf, p. 5
12
Michigan Department of State. (2013, January 4). Election results, general election, November 6, 2012. Retrieved
from http://miboecfr.nictusa.com/election/results/12GEN/90000003.html
13
Michigan Public Service Commission, op. cit., p. 4
Institute of Political Economy, Utah State University | 5
Commission (MPSC) stated that “Progress toward the 2012 compliance year and the 10 percent
renewable energy standard in 2015 is going smoothly.”14 It is estimated that by the end of 2015,
all but three of Michigan's seventy-two utilities will be on track to meet the RPS target.15
In response to the failure of Proposition 3 in 2012, the University of Michigan Energy Institute
studied how an expansion of the RES could solve Michigan’s future energy and environmental
concerns. The report examined multiple RES-expansion scenarios, and found that each scenario
bodes well for the Great Lake State, reducing reduced GHG emissions at little cost to electricity
ratepayers.16 The study’s findings have spurred new efforts to bolster the RES. In March of 2015,
Gov. Rick Snyder called on the Legislature to pass a slew of energy policies, mostly focusing on
reducing energy waste, but also mentioning the need for the state to derive at least 30 percent of
electricity from renewables by 2025. 17 Lawmakers responded by introducing six energy
efficiency and electricity industry reform bills—HB 4055, HB 4518, HB 4519, SB 295, SB 296,
and
SB
297—known
collectively
as
the
Powering
Michigan’s
Future
legislative
package. 18 , 19 , 20 ,21 , 22 , 23 HB 4518 and SB 295 are companion bills that establish a new RES
14
Ibid., p. 29
Ibid., p. 9
16
Johnson, J., Novacheck, J., Barteau, M., & Lyon, T. (2015, January). Expanding the Renewable Portfolio Standard
for Michigan: A study. University of Michigan Energy Institute. Retrieved from
http://energy.umich.edu/sites/default/files/michigan_rps_report_download_size_0.pdf
17
State of Michigan Executive Office. (2015, March 13). A special message from Gov. Rick Snyder: Ensuring
affordable, reliable, and environmentally protective energy for Michigan’s future. Retrieved from
https://www.michigan.gov/documents/150313_Energy_Message_FINAL_484033_7.pdf
18
Michigan Legislature. (n.d.). House Bill 4055. Retrieved from
http://www.legislature.mi.gov/(S(zbz0cp0rgfi5pegfps02grol))/mileg.aspx?page=GetObject&objectName=2015-HB4055
19
Michigan Legislature. (n.d.). House Bill 4518. Retrieved from
http://www.legislature.mi.gov/(S(4talskiof01myd3cb2jobdbn))/mileg.aspx?page=getobject&objectname=2015-HB4518
20
Michigan Legislature. (n.d.). House Bill 4519. Retrieved from
http://www.legislature.mi.gov/(S(4talskiof01myd3cb2jobdbn))/mileg.aspx?page=getObject&objectname=2015-HB4519
21
Michigan Legislature. (n.d.). Senate Bill 295. Retrieved from
http://www.legislature.mi.gov/(S(sqzpwbuky3y1fv5pagapmgns))/mileg.aspx?page=getobject&objectname=2015SB-0295
6
|
Renewable Portfolio Standards: Michigan
15
requirement of 20 percent renewable electricity by 2022, while the others focus on energyconservation and cost-containment.24,25
Some legislators, however, stand in opposition to a stronger RES. In March, Rep. Aric Nesbitt
introduced HB 4297, which would alter the RES to expand the qualifying source of renewable
electricity to include more methods of burning solid waste, and eliminate current energy
optimization standards in P.A. 295.26 Senator Mike Nofs is slated to introduce a bill that would
repeal both the energy optimization standards as well as the RES, focusing the state’s attention
instead on nuclear generation and improving emissions standards for traditional electricity
sources.27,28
In times when competing legislation would alter Michigan’s energy policies in such diametric
ways, citizens and legislators will be looking for information on the implications of the various
policy proposals. The conflict between these positions makes the need for accurate
measurements of the costs of RPS all the more pressing, and we hope the findings of our report
will inform officials on how best to serve their state.
22
Michigan Legislature. (n.d.). Senate Bill 296. Retrieved from
http://www.legislature.mi.gov/(S(mo51j5l4ropyt0xgkffwlgab))/mileg.aspx?page=getobject&objectname=2015-SB0296
23
Michigan Legislature. (n.d.). Senate Bill 297. Retrieved from
http://www.legislature.mi.gov/(S(sqzpwbuky3y1fv5pagapmgns))/mileg.aspx?page=getObject&objectname=2015SB-0297
24
Michigan Legislature, HB 4158, op. cit.
25
Michigan Legislature, SB 295, op. cit.
26
Michigan Legislature. (n.d.). House Bill 4297. Retrieved from
http://www.legislature.mi.gov/(S(irmgmlcehwq4hr3kbrnfcpht))/mileg.aspx?page=getobject&objectname=2015-HB4297
27
McArthur, C. (2015, May 7). Energy policy debate is missing the point, advocates say. Great Lakes Echo.
Retrieved from http://greatlakesecho.org/2015/05/07/energy-policy-debate-is-missing-the-point-advocates-say/
28
Balaskovitz, A. (2015, March 6). Michigan Republicans target RPS, efficiency standards. Retrieved from
http://www.midwestenergynews.com/2015/03/06/michigans-renewable-efficiency-standards-on-chopping-block/
Institute of Political Economy, Utah State University | 7
RESULTS
EMPIRICAL ANALYSIS
Analysis Performed by Tyler Brough, Ph.D., at Utah State University
STATE COINCIDENT EVENT STUDY
In this section, we present the results of an event study for state coincident indices—a
methodology first fashioned by the Federal Reserve Bank of Philadelphia.29 The event study
indexes the economic conditions of all states across multiple points in time, and assigns as “point
zero” each state’s economic conditions on the dates of their respective RPS implementations.
The study compares said economic conditions over a span from 48 months before to 48 months
after that enactment date. The indices of each state RPS policy, therefore, while enacted in
different calendar months and years, can thus be lined up in this so-called “event time,” and the
economic conditions in each state can be averaged. Given that RPS have been implemented in
many states over a long period, this will minimize the effects of anomalies such as recessions
and the enactment of other energy-related laws. For these reasons, the event study has become a
time-honored empirical methodology in finance and economics and a standard course of analysis
for the Philadelphia Fed. It is a simple but powerful method for measuring the effect of an
exogenous shock to an economic variable of interest. MacKinlay gives an in-depth discussion of
the event study methodology.30 Table 1 presents the dates of 31 different states that have enacted
an RPS policy.
29
Federal Reserve Bank of Philadelphia. (2015, January 29). State coincident indexes. Retrieved from
http://www.philadelphiafed.org/research-and-data/regional-economy/indexes/coincident/
30
MacKinlay, A.C. (1997). Event studies in economics and finance. Journal of Economic Literature, 35(1), 13-39.
8
|
Renewable Portfolio Standards: Michigan
TABLE 1: THE DATES (MONTH AND YEAR) OF THE 31 STATES THAT HAVE ENACTED AN RPS
POLICY TO DATE
State
Arizona
California
Colorado
Connecticut
Delaware
Hawaii
Iowa
Illinois
Kansas
Massachusetts
Maryland
Maine
Michigan
Minnesota
Missouri
Montana
North Carolina
New Hampshire
New Jersey
New Mexico
Nevada
New York
Ohio
Oregon
Pennsylvania
Rhode Island
South Carolina
Texas
Washington
Wisconsin
West Virginia
RPS Enactment Date
July, 2007
January, 2003
December, 2004
July, 1998
July, 2005
December, 2003
January, 1983
August, 2007
July, 2009
April, 2002
January, 2004
March, 2000
October, 2008
February, 2007
November, 2008
April, 2005
January, 2008
July, 2007
September, 2001
September, 2007
January, 1997
September, 2004
May, 2008
January, 2007
February, 2005
June, 2004
June, 2014
September, 1999
November, 2006
December, 2001
July, 2009
The results of the event study are presented in Figure 1, wherein we see the response of the state
coincident index to the enactment of RPS policies. The coincident index is a summary measure
Institute of Political Economy, Utah State University
|
9
of the strength of a state economy, and is comprised of four economic indicators: nonfarm
payroll employment, average hours worked in manufacturing, the unemployment rate, and wage
and salary disbursements deflated by the Consumer Price Index (CPI).31
31
Federal Reserve Bank of Philadelphia. (2015, January 29). State Coincident Indexes. Retrieved from
http://www.philadelphiafed.org/research-and-data/regional-economy/indexes/coincident/
10
|
Renewable Portfolio Standards: Michigan
FIGURE 1: THE RESPONSE OF THE STATE COINCIDENT INDEX TO THE ENACTMENT OF RPS
POLICIES.
The horizontal axis shows months before and after point zero (RPS enactment). The vertical axis shows an indexed
scale measuring the average reaction of states in terms of several economic indicators.
As can be seen in Figure 1, the average effect on the state coincident index is a precipitous drop
surrounding the enactment of an RPS policy. This evidence is suggestive of a negative effect of
an RPS policy on a state economy. While suggestive, the evidence from the event study warrants
further exploration into the effects, since state economies also appear to decline several months
prior to the enactment of an RPS. The next section presents the structural panel VAR-X model,
which provides further evidence of the negative economic effects of an RPS.
Institute of Political Economy, Utah State University
|
11
THE STRUCTURAL PANEL VAR-X MODEL
The VAR model takes into account the nature of the state macroeconomic variables that could
provide unwanted feedback into the model, and considers their dynamic interactions. By
including a panel dimension to the model we can include the data for multiple states in a single
model. We include fixed effects to control for state-level heterogeneity. We impose a recursive
causal ordering on the VAR-X model to allow for structural interpretation of dynamic multiplier
analysis of the RPS policy variable. Table 2 presents the cumulative effects of an RPS on the
state economy via structural policy simulations.
TABLE 2: THE LONG-RUN EFFECTS ON STATE MACROECONOMIC VARIABLES
State Economic Variable
Electricity Sales
Real Personal Income
Non-farm Employment
Manufacturing Employment
Unemployment Rate
Long-Run Effect
-13.7075%
-3.6369%
-2.8416%
3.7454%
9.6841%
The cumulative effect of the enactment of an RPS policy on state electricity sales is a staggering
13.7-percent decline. This is, perhaps, not surprising as the RPS increases the cost of electricity
generation. Real personal income declines in the long run by 3.6369 percent, which figures to a
loss of $15.1 billion in 2013, or $3830 less per family.32 Non-farm employment declines in the
long run by 2.8 percent. Only one analyzed component of non-farm employment, manufacturing
employment, does not experience a long-term suppression in response to an RPS policy,
although as we see in the later graphical analysis, it does still experience a sharp decline in the
32
Bureau of Economic Analysis. (n.d.). Regional Data, Annual State Personal Income and Employment. Retrieved
from http://www.bea.gov/iTable/iTable.cfm?reqid=70&step=1&isuri=1&acrdn=1#reqid=70&step=1&isuri=1
12
|
Renewable Portfolio Standards: Michigan
short term. Most significantly, the state unemployment rate increases by 9.6 percent. This means
that, in March of this year, Michigan had 24,369 fewer jobs than it would have had without the
RPS. 33 There can be no doubt that the combined economic effect on an RPS enactment, as
measured by the structural panel VAR-X model, is a severe decline in the Michigan economy. A
graphical representation of the analysis, showing the changes over time that lead to these results,
can be found in Appendix B.
CONCLUSIONS FROM THE EMPIRICAL ANALYSIS
We demonstrate strong empirical evidence that a Renewable Portfolio Standard has a lasting
negative effect on a state economy. We present this evidence from both an event study of the
state coincident index as measured by the Federal Reserve Bank of Philadelphia, as well as from
structural policy simulations from a panel VAR-X model. The long-run effect of an RPS on state
industrial production, as measured through industrial electricity sales, is almost a 14-percent
decline. Real personal income declines in the long run after an RPS by almost four percent. The
cumulative effect of an RPS on non-farm employment is nearly three percent. While the effect of
an RPS on manufacturing employment is not as severe in the long run, it too demonstrates an
initial sharp decline lasting for several years. Finally, the state unemployment rate increases in
the long run in response to an RPS by nearly 10 percent. These are strong and lasting effects in
four of the five variables measuring the state economy. The combined econometric evidence
makes clear that an RPS policy has a severely negative economic effect on a state that enacts
such.
33
Bureau of Labor Statistics (n.d.). Michigan. Retrieved from http://www.bls.gov/regions/midwest/michigan.htm
Institute of Political Economy, Utah State University | 13
INSTITUTIONAL ANALYSIS
Our survey of legal rules, or “institutional analysis,” explores how the regulatory structure of
Michigan’s Renewable Energy Standard (RES) creates barriers to compliance for state electricity
providers, municipal utilities, investor-owned utilities, rural electric cooperatives, and retail
electricity suppliers. We find that even if RPS posed no economic harms, which our economic
analysis disputes, the overly complex and contradictory structure of Michigan’s RES will
inherently hinder compliance.
As stated earlier, Michigan’s RES requires all electric providers to supply at least ten percent of
their total retail sales using renewable energy resources by 2015.34 Additionally, the two largest
investor-owned electricity companies, Detroit Edison and Consumers Energy, must meet an
additional capacity requirement of 500 and 600 megawatts, respectively, by the end of 2015.35
There is also a capacity requirement for other large-scale providers: If a provider had 1,000,000
to 2,000,000 retail electric customers as of January 1st, 2008, it must produce 200 megawatts
(MW) of renewable electricity by the end of 2013, and 500 MW by the end of 2015. If a provider
had more than 2,000,000 retail electric customers, it must produce 300 MW by the end of 2013
and 600 MW by the end of 2015.36
In Michigan there are currently no minimum technology requirements in place. However, for
hydroelectric facilities renewable energy generation is only eligible where the dam is a repair or
replacement of one existing before October 2008, or the dam is upgraded to increase energy
efficiency. For municipal solid waste incineration, renewable energy generation is only eligible
at existing facilities where efficiency is increased or limited expansion occurs. Only allowing
34
United States Department of Energy. (2014). Database of state incentives for renewables & efficiency. Retrieved
from http://programs.dsireusa.org/system/program/detail/3094
35
Ibid.
36
Ibid.
14
|
Renewable Portfolio Standards: Michigan
utilities to improve or repair their facilities, and not create new ones, can hinder compliance now
and will surely hinder it in the future.
An important component of the RES is the complex incentive system involving compliance
multipliers. Michigan’s RES does not mandate any “carve-outs” for particular generation
sources, so it instead incentivizes particular sources by compounding value of their attendant
RECs. For example:37
•! For each megawatt hour (MWh) produced from solar power, two additional RECs are
earned.
•! For each MWh generated from a renewable energy system other than wind, at on-peak
demand time,38 an additional 0.2 RECs are earned.
•! For each MWh generated from a renewable energy system during off-peak hours, using
advanced electric storage technology; or a hydroelectric pumped storage facility during
on-peak hours,39 an additional 0.2 RECs are earned.
Because labor and capital generally flows to higher-value uses, these multipliers are not
necessary, provided that there is market demand for electricity generated from the governmentfavored sources. If there is insufficient market demand for electricity generated from those
sources then incentivizing said generation will lead to inefficiency, and lower consumer surplus.
The RES also uses multipliers to support Michigan’s home manufacturing base. For example:40
•! For each MWh generated from a renewable energy system constructed using equipment
made in Michigan, an additional 0.1 RECs are earned.
37
Ibid.
On-peak times include weekdays at a chosen time period, for example 7:00 AM to 7:00 PM.
39
Off-peak times are early mornings, nights, weekends, and holidays.
40
United States Department of Energy (2014), op. cit.
Institute of Political Economy, Utah State University
38
|
15
•! For each MWh produced from a system constructed using a workforce composed of
Michigan residents, an additional 0.1 RECs are earned.
Unfortunately for Michigan manufacturers, these multipliers apply only for the first three years
after the system begins commercially producing electricity. In the case of the first multiplier, this
may cause utilities to purchase exclusively from Michigan manufacturing companies but switch
to a cheaper option once the three-year REC multiplier ceases, pushing Michigan manufacturers
out of business. The large in-state manufacturers will be able to weather this storm, but small
companies may not be able to compete in the absence of artificial support. In the case of the
second multiplier, the artificially high amount of electricity purchases from facilities using
equipment made by domestic employees may cause greater hiring in the short term, but cause a
large round of layoffs three years down the road. In another scenario, business owners may
anticipate these changes and refrain from producing more energy equipment or hiring more
workers in the first place, rendering the multipliers self-defeating.
Utilities buy and sell RECs to assist in meeting their renewable energy goals.41 The Michigan
Public Service Commission (MPSC) administers the REC market through the Michigan
Renewable Energy Certification System (MIRECS), a certification and tracing program which
monitors the trading, selling, and transferring of RECs.42 One important stipulation of RECs in
Michigan, however, is that if the owner does not sell an REC produced at his facility within three
years, then the REC may no longer be sold on MIRECS. While utilities have accrued RECs in
excess of what they need to satisfy the RES up to 2015, most of these RECs were generated in
2010, and will be useless of the Legislature sets a new RES requirement for the future.43 As of
2014, MPSC reported that 2.9 million RECs had expired without being used for compliance.44
41
Michigan Public Service Commission, op. cit., p. 12
Ibid., p. 12
43
Ibid., p. 7
44
Ibid.
16
|
Renewable Portfolio Standards: Michigan
42
The regulatory uncertainty may have induced overproduction in the short term, forcing the
construction of otherwise unnecessary new generation facilities later on.
Utility companies are allowed to recover, through limited surcharges, their costs above and
beyond what they would have incurred without a RPS. For residential consumers, the surcharge
to recover these “incremental” costs cannot exceed $3.00 per month; for commercial consumers
the surcharge cap $16.58 per month; and for industrial consumers the cap is $187.50 per
month.45 Cost caps protect politicians and regulators from the ire of ratepayers, but in most states
they hinder the very achievement of the RPS. In Michigan, cost caps are not currently hindering
compliance—in fact the few utilities that impose surcharges have been lowering them—but that
could change if the Powering Michigan’s Future legislative package passes. This set of bills
would not only create a new RES requirement of 20 percent renewable electricity by 2022,46,47
but it would also ban cost-recovery surcharges entirely.48 Indeed, one of the scenarios in the
aforementioned University of Michigan Energy Institute study is a 20-percent RES by 2030, and
according to the Institute this would already translate into a per-household rate impact of $1.70
every month between now and 2035. 49 If legislators insist on the tougher RES in Powering
Michigan’s Future, and ban surcharges, that would place utilities in an impossible bind that may
prevent compliance.
Also, in order for rate-regulated utilities to recover incremental costs, the MPSC must first
approve those costs. In order to gain approval, said costs must be “reasonably and prudently
45
Michigan Legislature. (2008). Michigan Compiled Laws, Chapter 460, Act 295, Part 2, Subpart A, Section 1045,
Clean, Renewable, and Efficient Energy Act (Excerpt). Retrieved from
http://www.legislature.mi.gov/(S(iq2024pqot2pwqty5sz2d1m4))/mileg.aspx?page=getObject&objectName=mcl460-1045
46
Michigan Legislature, HB 4518, op. cit.
47
Michigan Legislature, SB 295, op. cit.
48
Michigan Legislature, SB 297, op. cit.
49
Johnson, Novacheck, Barteau, & Lyon, op. cit.
Institute of Political Economy, Utah State University | 17
incurred in good faith.” 50 The term “good faith” is invites subjectivity, however, and as the
Climate Policy Initiative notes, these qualifying costs are “often set at a level determined
politically, rather than based on expected costs.”51 Requiring the MPSC to consider all costs,
without requiring that they quantitatively measure the costs, does not provide substantial and
accurate information to utilities, but does slow down the process of reaching compliance. The
MPSC is also required to provide the legislature with an annual report on the implementation of
the state’s RES cost and effectiveness, however, they are not required their methods for
determining the impacts.52
If the MPSC finds that a utility did not make a good-faith effort to comply with the RES, then the
provider will not be able recover any surcharges it placed on ratepayers in order to purchase
RECs.53 If an alternative electric supplier is guilty of the same, the MPSC may revoke its license,
issue a cease and desist order, or enforce a fee of between $5,000 and $50,000.54 Lastly, if a
municipally-owned electric utility fails to meet the RES, the MPSC will notify the Attorney
General, who may commence civil action against the utility.55
50
Michigan Legislature. (2008, October 6). Enrolled senate bill no. 213. Retrieved from
http://www.legislature.mi.gov/documents/2007-2008/publicact/pdf/2008-PA-0295.pdf
51
Pierpont, B. (2012, June). Limiting the cost of renewables: Lessons for California. Retrieved from
http://climatepolicyinitiative.org/wp-content/uploads/2012/06/Limiting-the-Cost-of-Renewables-Lessons-forCalifornia.pdf
52
Michigan Legislature. (2008). Michigan Compiled Laws, Chapter 460, Act 295, Part 2, Subpart A, Section 1051,
Clean, Renewable, and Efficient Energy Act (Excerpt). Retrieved from
http://www.legislature.mi.gov/(S(2hlt3wpwrk1vrtqrfoilvp2e))/mileg.aspx?page=getobject&objectname=mcl-4601051&query=on&highlight=report
53
Michigan Legislature. (2008). Michigan Compiled Laws, Chapter 460, Act 295, Part 2, Subpart A, Section 1047,
Clean, Renewable, and Efficient Energy Act (Excerpt). Retrieved from
http://www.legislature.mi.gov/(S(2hlt3wpwrk1vrtqrfoilvp2e))/mileg.aspx?page=getObject&objectName=mcl-4601047
54
Ibid.
55
Michigan Legislature. (2008). Michigan Compiled Laws, Chapter 460, Act 295, Part 2, Subpart A, Section 1053,
Clean, Renewable, and Efficient Energy Act (Excerpt). Retrieved from
http://www.legislature.mi.gov/(S(2hlt3wpwrk1vrtqrfoilvp2e))/mileg.aspx?page=getobject&objectname=mcl-4601053&query=on&highlight=attorney%20AND%20general
18
|
Renewable Portfolio Standards: Michigan
Exceptions to meeting the requirement are granted for numerous reasons including: issues with
renewable energy system site requirements (if the providers exercised “reasonable diligence”),
excessive equipment cost or component shortages, issues with electric transmission and
interconnection, projected or actually unfavorable reliability or operational impacts, labor
shortages, and orders of court that limit the availability of renewable energy systems. 56 With so
many available exceptions, utilities may be able to spend less money fitting one of the exceptions
than actually implementing a renewable energy system.
CONCLUSIONS FROM THE INSTITUTIONAL ANALYSIS
The institutional analysis reveals that P.A. 295, and the RES specifically, contain some features
that are self-defeating in the realm of attaining compliance. REC multipliers hobble efficient
generation even among the various renewable electricity sources, and cost caps could stymie
compliance if legislators pass the Powering Michigan legislative package. This analysis
evaluates the RES on its own terms, and policymakers should not take for granted that the RES
will achieve its goals, even if we were to accept the economic costs.
CONCLUSION
The results from these analyses paint a clear picture about the effects of RPS. Our empirical
analysis finds a drop in industrial electricity sales by almost 14 percent, real personal income by
almost four percent, non-farm employment by nearly three percent and an increase of 10 percent
in the unemployment rate, which for Michigan equals a loss of 29,366 jobs. Our analysis of the
legal rules surrounding Michigan’s RES describes the barriers that make it difficult for utilities to
comply and for bureaucracies to enforce it, finding that Michigan’s multipliers and cost caps
56
Michigan Legislature. (2008). Michigan Compiled Laws, Chapter 460, Act 295, Part 2, Subpart A, Section 1031,
Clean, Renewable, and Efficient Energy Act (Excerpt). Retrieved from
http://www.legislature.mi.gov/(S(2hlt3wpwrk1vrtqrfoilvp2e))/mileg.aspx?page=getObject&objectName=mcl-4601031&highlight=reasonable%20AND%20diligence
Institute of Political Economy, Utah State University | 19
hinder compliance by creating inefficiency and potentially erecting political barriers to satisfying
the mandates. Any state currently deliberating on implementing a new RPS, or strengthening an
existing one, should heed these results as a warning of their harmful effects. Finally, states
should refrain from following the fad of enacting such costly regulations, in spite of the policy’s
political palpability or expediency.
20
|
Renewable Portfolio Standards: Michigan
APPENDIX A
EXPLANATION OF EMPIRICAL STUDY METHODOLOGY
Methodology Constructed by Tyler Brough, Ph.D. at Utah State University
In this technical appendix, we outline the details of the structural panel VAR-X model, its
estimation, and its use for policy simulation.
0.1. THE PANEL VAR-X MODEL
The vector autoregressive (VAR) model is the standard work horse model in empirical
macroeconomics. The basic p-lag VAR model can be written as:
)
!" = $0 +
'( !",( + -"
(*+
where !" for t = 1, . . . , T is an M vector of observations on M time series variables, -" is an M ×
1 vector of errors, $. is an M × 1 vector of intercepts and the '( are M × M matrices containing
model coefficients. 57 This is the reduced-form VAR model. For the present study !" =
(!+" ,...,!/" )′, M = 5, and the !(" are the five state macroeconomic variables presented in the main
body of the paper, namely electricity sales, real personal income, non-farm employment,
manufacturing employment, and the unemployment rate. Thus, the VAR model is a system of M
equations, with one equation for each variable in the system. Each of the M = 5 variables is
treated as endogenously determined.
57
Lutkepohl, H. (2005, Spring). New introduction to multiple time series analysis, Chapter 2.
Institute of Political Economy, Utah State University
|
21
The present model also includes an exogenous policy variable that represents the enactment of an
RPS by a given state. Thus, we can now write the VAR-X (a VAR model with the exogenously
determined variable) as follows:
)
2
!" = $. +
'( !",( +
(*+
0",1 3",1 + -"
1*+
where the 3",1 vectors contain the exogenous variables and their lags, and the B matrices
contain the coefficients respectively. The variables in the X vector affect the state of the other
variables, but are not themselves determined by the system of equations, and thus are considered
exogenous.
In addition, the present model adds a cross-sectional dimension to the basic VAR(p) model by
essentially stacking the VAR models for the different states on top of each other. In other words,
the panel VAR-X model is:
)
2
!+," = $+,. +
'( !+,",( +
(*+
0",1 3",1 + -+,"
1*+
⋮
)
!/," = $/,. +
2
'( !/,",( +
(*+
0",1 3",1 + -6, 7
1*+
See Canova and Ciccarelli for an excellent survey of panel VAR methods.58 One thing of note is
that in the present study we do not focus on estimating dynamic heterogeneities between the
different state economies as is done in many panel VAR models for large macroeconomic
studies. Instead we focus on the average effect of an RPS enactment on a state economy. To that
58
Canova, F., & Ciccarelli, M. (2013). Panel vector autoregressive models: A survey.
22
|
Renewable Portfolio Standards: Michigan
end, we estimate the model with fixed effects to control for possible heterogeneities across
states.59 It is possible to recover the state fixed effects, though we make no effort to do so here as
the focus of the study is on the average effect of an RPS and not on individual state effects.
0.2. MODEL ESTIMATION
We use Bayesian techniques, namely the Gibbs sampler, to estimate the structural panel VAR-X
model. See Ciccarelli and Rebucci for a review of Bayesian methods for VAR models.60 See also
Ocampo and Rodríguez for a very practical tutorial.61 We follow their Algorithm 3 for Bayesian
estimation, which for simplicity we reproduce below.
Algorithm 1 Bayesian Estimation
1.! Select the specification for the reduced form VAR-X, that is to choose values of p
(endogenous variables lags) and q (exogenous variables lags) such that the residuals of
the VAR-X ( - ) have white noise properties. With this the following variables are
obtained: T, p, q, k, where:
8 = 1 + :; + <(> + 1)
2.! Calculate the values of Γ, S with the data (Y , Z) as:
Γ = (A B A),+ A′D
E = (D − AΓ)′(D − AΓ)
3.! Generate a draw for matrix Σ from an inverse Wishart distribution with parameter S and T
− k degrees of freedom.
Σ~IJ)KL (E, M − 8)
59
Greene, William. (2012). Econometric Analysis. Upper Saddle River, NJ: Prentice Hall.
Ciccarelli, M., & Rebucci, A. (2003). Bayesian VARs: A survey of recent literature with an application to the
European monetary system.
61
Ocampo, S., & Rodríguez, N. (2012). An introductory review of a structural VAR-X estimation and applications.
Revista Colombiana de Estadística, 3, 479-508.
Institute of Political Economy, Utah State University | 23
60
4.! Generate a draw for matrix Γ from a multivariate normal distribution with mean Γ and
covariance matrix ΣN (A B A),+
Γ|Σ~6P)KL (Γ, Σ⨂ A B A
,+
)
5.! Repeat steps 2-3 as many times as desired, save the values of each draw. The draws generated can be used to compute moments of the parameters. For every draw the
corresponding structural parameters, impulse response functions, etc. can be computed, then
their moments and statistics can also be computed. The algorithm for generating draws for the
inverse Wishart and multivariate normal distributions are presented in Bauwens et al., Appendix
B.62
Observe that in this notation:
!
#
#
#
Y =#
#
#
#"
y'1 $
&
! &
&
y't &
! &
&
y'T &%
,
62
Bauwens, L., Lubrano, M., & Jean-Francois, R. (2000). Bayesian inference in dynamic econometric models.
Oxford, UK: Oxford University Press.
24
|
Renewable Portfolio Standards: Michigan
"
$
$
$
Z =$
$
$
$
#
1
y'0
!
y'1− p
"
1
y't−1
!
y't− p
"
1
y'T −1
!
y'T − p
x '1 %
'
'
'
x 't '
'
'
x 'T '
&
,
!
#
#
#
E =#
#
#
#"
ε '1 $
&
! &
&
ε 't &
! &
&
ε 'T &%
and finally,
Γ = "$ v
#
A1
…
Ap
B0
!
Bq %'
&
Then the VAR-X model can be written simple as
D = AΓ + R
We set ; = 3 and > = 1 for simplicity.
Institute of Political Economy, Utah State University
|
25
0.3. DYNAMIC MULTIPLIER ANALYSIS
During the Gibbs sampling simulation, which we run for 5, 000 replications with 500 burn-in
steps, we also conduct dynamic multiplier analysis for the exogenous RPS policy variable. We
follow Algorithm 2 in Ocampo and Rodrıguez to conduct this analysis.63 This algorithm is as
follows:
Algorithm 2 Identification by Long-Run Restrictions
1.! Estimate the reduced form of the VAR-X model. 2.! Calculate the VMA-X representation of the model (matrices TU ) and the convariance
matrix of the reduced form disturbance - (matrix Ʃ).
3.! From the Cholesky decomposition of T(1) and ƩT(1) calculate matrix C(1)
V 1 = chol(T 1 ƩTB 1 )
4.! With the matrices of long run effects of the reduced form, \(1), and structural shocks,
C(1), calculate the matrix of contemporaneous effects of the structural shocks, V. .
V. = T 1
,+
V(1)
5.! For i = 1, . . . , R with R sufficiently large, calculate the matrices VU as:
VU = TU V.
Identification is completed since all matrices of the structural VMA-X are known.
63
Ocampo, S., & Rodríguez, N., op. cit.
26
|
Renewable Portfolio Standards: Michigan
We set R = 120 months after an RPS to estimate the cumulative, or long-run effects of an RPS
enactment for dynamic multiplier analysis.
Institute of Political Economy, Utah State University
|
27
APPENDIX B
GRAPHICAL
ANALYSIS
OF
THE
DYNAMIC
MULTIPLIER
ANALYSIS
Below we present in graphical form the dynamic multiplier analysis for each of the five state
macroeconomic variables. This analysis strengthens the evidence of a severely deleterious effect
of an RPS policy. For electricity sales, real personal income, and non-farm employment the
response to an RPS is an initial sharp decline lasting for several years and the long-run effect is a
large and lasting decline. Manufacturing demonstrates the same initial sharp decline in response
to an enacted RPS, but does show some recovery, after several years, though still never returns to
levels prior to the RPS. However, the unemployment rate demonstrates a steadily increasing rate
that cumulates into a large increase in state unemployment.
Dynamic Multiplier Analysis for Electricity Sales.
28
|
Renewable Portfolio Standards: Michigan
Dynamic Multiplier Analysis for Real Personal Income.
Dynamic Multiplier Analysis for Non-farm Employment.
Institute of Political Economy, Utah State University
|
29
Dynamic Multiplier Analysis for Manufacturing Employment.
Dynamic Multiplier Analysis for the Unemployment Rate.
30
|
Renewable Portfolio Standards: Michigan
PRINCIPAL INVESTIGATORS
RANDY T SIMMONS PH.D
UTAH STATE UNIVERSITY
RYAN M. YONK PH.D
UTAH STATE UNIVERSITY
TYLER BROUGH PH.D
UTAH STATE UNIVERSITY
KEN SIM
STRATA POLICY
JACOB FISHBECK
STRATA POLICY