Evaluating the Effectiveness of the Community Rating System: A Comparative Analysis (and other CRS related tidbits) Wesley E. Highfield & Samuel D. Brody Center for Texas Beaches & Shores Department of Marine Sciences Texas A&M University at Galveston
Previous Nationwide Analysis What activities standout as effective? How well does the program perform nationwide? Track CRS communities over a 11 year period: 1999 2009. Panel regression models CRS activity points and other factors (control variables) Isolate the effect of each selected CRS activity by controlling for other pertinent variables
U.S. Insured Flood Losses: 1999 2009 (Millions of 2012 U.S. Dollars) Nationwide: $33.5 Billion in paid claims 90.00% Percent of Insured Flood Damage in 1% Flood Zones 24% 80.00% 70.00% 60.00% 3% 50.00% 40.00% 73% 30.00% A Zone V Zone SFHA OUT 20.00% 10.00% 0.00% 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Previous Nationwide Analysis Nationally representative sample of 450 CRS communities Unit of Analysis is community Losses and predictors are aggregate Cross sectional time series (CSTS)/panel models» Linear, Random Effects Panel Regression Models Quantify in dollar amounts effect of CRS activities on insured flood loss claims
Previous Nationwide Analysis Open space (420) Freeboard (element of 430) Flood Protection (530) Average Loss Reduction per CRS community Building and Contents Content Damage Building Damage Total A V B C D X Total A V B C D X Total A V B C D X 420 $547,497 $483,869 $95,441 $56,482 $444,715 $399,146 FRB $960,817 $669,260 $96,200 $70,927 $780,441 $537,547 $44,640 530 $416,016 $327,575
Drilling Down: CRS in a Texas Gulf Coast Watershed National level important programmatically What works one place may not work in another Analysis at the community level can wash out some important details Structural and geographic characteristics What is the effect of the CRS program and its activities relative to non CRS communities on insured losses?
Clear Creek Study Area 197 square miles Adjacent to Galveston Bay Portions of four counties 22 communities o 12 CRS o 10 Non CRS Study period: 1999 2009
Clear Creek Analysis: Approach Track NFIP claims and CRS activity points from 1999 2009 Statistical controls included: Elevation (ground) Precipitation Home Age Home Value Floodplain Proximity Post FIRM status (dichotomous) Year All measurements/unit of analysis at the parcel level Analyzed with spatial autoregressive (SAR) models Spatial autocorrelation in the error term Overall sample: n = 9,555 claims over study period
Clear Creek: Descriptive Results Over $335 million in claims from 1999 2009 46% of claims outside of the SFHA 42% of claims were for Pre FIRM properties Average per property claim approximately $35K TS Allison (2001) generated 40% of claims Hurricane Ike (2008) generated 46% of claims Mean insured losses by year: 1999 2009
Clear Creek Analysis: Results Control variables behave as expected Control Variable Direction Elevation Precipitation + Home Age + Home Value + Floodplain Proximity Post FIRM Property On average: 1 increase in ground elevation decreases damage by $5,985 $7,785 100 away from the SFHA decreases damage by 4 5% Post FIRM properties 79 86% less damage
Clear Creek Analysis: CRS Results PARTICIPATION: On average, structures in CRS communities saw 88% reduction in claim amount Relative to structures in non CRS communities» After controlling for previous variables CRS POINTS: One point increase in total CRS points reduced loss amount by 0.06% Seemingly small amount, but recall that classes move in 500 point increments. CRS ACTIVITIES: All CRS activities analyzed had statistically significant effects on reducing loss amounts in the Clear Creek Watershed.
Clear Creek Analysis: CRS Results Activity Mean Points Max Possible Mean Per Point Reduction Total Mean Reduction 320: Map Information 124 140 $177 $16,240 330: Outreach Projects 110 315 $205 $16,457 340: Hazard Disclosure 12 81 $358 $4,105 350: Flood Protection Information 32 66 $845 $18,544 360: Flood Protection Assistance 33 71 $310 $8,893 410: Additional Flood Data 29 1,373 $443 $10,841 420: Open Space Protection 106 900 $82 $7,671 430: Higher Regulatory Standards 259 2,720 $126 $21,023 440: Flood Data Maintenance 90 231 $395 $22,133 450: Stormwater Management 69 670 $187 $10,764 510: Floodplain Management Planning 64 309 $344 $16,187 520: Acquisition/Relocation 317 3,200 $41 $10,792 540: Drainage System Maintenance 216 330 $143 $20,302 610: Flood Warning Program 84 225 $109 $8,071
Clear Creek Analysis: CRS Results $25,000 Mean Loss Reduction by Activity $20,000 $15,000 $10,000 $5,000 $0 320 330 340 350 360 410 420 430 440 450 510 520 540 610 Activities 440 (flood data maintenance), 430 (higher regulatory standards) and 540 (drainage system maintenance) had the largest effect. Followed closely by 510 (FP management planning), 350 (flood protection information), 330 (outreach projects), and 320 (map information) BUT: is a function of how many points communities have (or have not) accrued in each activity
Clear Creek Analysis: CRS Results 100% Percent of Potential Points Accrued by Activity 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 320 330 340 350 360 410 420 430 440 450 510 520 540 610 Still lots of room to move, esp. in 430.
Conclusions from a smaller scale When comparing damage between CRS/Non CRS communities: Homes in CRS participating communities had significantly lower claim amounts Individual CRS Activities significantly reduce loss amounts BUT: Variation in activity effectiveness Differential between per point and mean savings Lots of room to improve based on mean points and maximum points The right mix is dependent on scale/situation
Moving Forward Individual level analyses are nice allow measurements that cannot be done at community level Can still suffer from a lack of variation across CRS variables Currently scaling back up to community to conduct comparative, nation wide analysis of CRS vs. non CRS (NFIP) communities
Comparing CRS vs non CRS in the U.S. Approach must be different 20,000+ NFIP communities; < 1200 CRS communities Currently in the process of matching CRS and non CRS communities. Propensity Score Matching Current matching criteria include: Property Tax Rate ( ) Education (+) Housing Units (+) Proportion of SFHA (+) Structures in SFHA ( ) NFIP Policies (+) Year Built (+) NFIP Entry Year ( ) Coastal (+) Precipitation (*)
Comparing CRS vs non CRS in the U.S. Overarching goal, after establishing the matched communities: Determine the average effect of CRS program on flood loss claims Relative to non CRS communities
Thanks Questions? Highfield, W.E., Brody, S.D. Evaluating the Effectiveness of Local Mitigation Actions in Reducing Flood Losses, Natural Hazards Review, In Press. Brody, S.D., Highfield, W.E., Open Space Protection and Flood Mitigation: A National Study, Land Use Policy, In Press Highfield, W.E., Brody, S.D. and Blessing, R. Measuring the Impact of Mitigation Activities on Flood Loss Reduction at the Parcel Level: The case of the Clear Creek watershed on the upper Texas coast, Natural Hazards, In Press.