Asking the Right Que stions to Make Data Dr ive n De c isions on - PowerPoint PPT Presentation

Asking the Right Que stions to Make Data Dr ive n De c isions on Staffing and De ployme nt E xc lusive Pr o vide r o f Public Safe ty T e c hnic al Se r vic e s fo r Inte r natio nal City/ Co unty Manage me nt Asso c iatio n 1

L e o na rd Ma ta re se , Dire c to r o f Re se a rc h Ce nte r fo r Public Safe ty Manage me nt, L L C L ma ta re se @ c psm.us • 716-969-1360 2

Q. Why do you r ob banks? 3 3

“Be c ause that’s whe r e the mone y is.” – Willie Sutto n 4 4

T he r e r e ally is a “Sutton’s L aw”: • In me dic ine state s that whe n diagnosing, one should fir st c onside r the obvious; • Use d in g (ABC) of manage me nt ac c ounting, the law stipulate s that ABC should be applie d "whe r e the mone y is," me aning whe r e the highe st c osts ar e inc ur r e d and, thus, the highe st pote ntial of ove r all c ost r e duc tion is. 5 5

Ove r vie w • T ypic al Re que st is MORE r e sour c e s PL E ASE ! – L imite d analyse s of wor kloads, pe r for manc e , and value of MORE ? • Aggr e gate data on total c alls misle ading • Pe r for manc e NOT line ar func tion of total c alls: – An inc r e ase of 50% may r e quir e little or no additional r e sour c e s 6

Polic e Issue s 7

Polic e Patr ol: Re ac tive and Pr oac tive Policies & Process Management Procedures Citizen Expectation Continuous Citizen Complaints Improvement Citizen Satisfaction Response Time Reduction Feedback Data Driven Loop Targeted Crime Reduction Citizen Fear Community Deterioration Leverage Resources Resource Flexible and Creative Math Model Allocation 8

Q. What is the r e c omme nde d offic e r s pe r thousand for a jur isdic tion of your size ? 9

A. T HE RE IS NO SUCH T HING! 10

No T wo Jur isdic tions ar e the Same • Ge o g ra phy • So c ia l – E c o no mic Co nditio ns • Crime Ra te • Physic a l Size • De nsity 11

De ployme nt Que stions to Ask Q. How ma ny pa trol pe rsonne l a re a c tua lly on the stre e t? By Se a son, Da y of We e k, Hour of da y? Q. Wha t’s the diffe re nc e be twe e n Ca lls F or Se rvic e (CF S) vs. Workloa d – CF S is a quantity – a numbe r – Wor kload is the c umulative time to handle CF S 12

E xample of CF S vs. Wor kload • Bur glar y – we se nd two offic e r s to handle c all. T he y ar e the r e for 1 hour e ac h. • Calls for Se r vic e = 1 • Wor kload = 2 Hour s (2 x 1 hour s) 13

Re sponse T ime s Que stions Q. Do you know that r e sponse time is only impor tant on limite d numbe r of high pr ior ity c alls? Q. Ar e c alls dispatc he d and r e c or de d by pr ior ity? Q. Is r e sponse time me asur e d by time of day, day of we e k and se ason? Q. Do the y always c he c k for bad data? Q. Do the y tr ac k long r e sponse time s for high pr ior ity c alls and find out why? 14

Ke y Patr ol Que stions Q. Is de ployme nt aligne d with wor kload var iations by se ason, day of we e k, hour of day? Q. Ar e same numbe r of patr ol units assigne d ar ound the c loc k? – r ar e ly good ide a Q. Ar e pr oble m ar e as suc h as shift c hange ide ntifie d? Q. Do shift star t time s, sc he dule s & CBAs limit de ployme nt fle xibility and alignme nt? 15

Polic e - City of 275,000 3 star t time s/ 10- hour ove r lapping shifts 0200 to 0700 to 1600 to 2100 to 0700 1600 2100 0200 August 2002 29 33 33 62 Januar y 2003 32 34 32 64 Da ily Ave ra g e Ca ll Ra te  August 34% highe r than Januar y  Pr ior ity 1 & 2: 50% mor e c alls pe r hour in August 16

Misa lig ne d Pa trol F orc e - City of 275,000 3 sta rt- time s/ 10- hour shifts 80 Deployed Busy at Calls 60 Patrol Units 40 20 0 00 02 04 06 08 10 12 14 16 18 20 22 Time of Day 17

Alig ne d Pa trol F orc e – City of 275,000 5 sta rt- time s/ 10- hour shifts Deployed 100 Busy at Calls 80 Patrol Units 60 40 20 0 00 02 04 06 08 10 12 14 16 18 20 22 Time of Day 18 18

T he Polic e F oundation Shift L e ngth E xpe r ime nt 19

Shift L e ngth E xpe r ime nt L imitations: • Se lf Re porting • Not Ra ndom Sa mple – Offic e rs se le c te d shifts. • Offic e rs we re a llowe d to c ha ng e shift in mid proc e ss. • L imite d to 3 De pa rtme nts 20

Shift L e ngth E xpe r ime nt Positive s: • E xc e lle nt bio da ta te sting 21

10- hour shifts ha ve se lf re porte d a dva nta g e s ove r 8- hour shifts • Offic e r s wor king 10- hour shifts got mor e sle e p pe r night (ove r half an hour ) than those on 8- hour shifts and had a signific antly highe r quality of wor k life (se lf r e por te d). • 10- hour shifts wor ke d the le ast amount of ove r time of the thr e e gr oups, pote ntially r e sulting in c ost savings. 22

T he be ne fits of 10- hour shifts ma y not e xte nd to 12- hour shifts • 10- hour shifts got mor e sle e p than those on 8- hour shifts; not tr ue for 12- hour shifts. • 10- hour shifts had a highe r r e por te d quality of wor k life than those on 8- hour shifts; 12- hour shifts did not. • 12- hour shifts wor ke d a le sse r amount of ove r time than those on 8- hour shifts; the y still wor ke d mor e than those on 10- hour shifts. 23

Shift le ngth impac t • Shift le ngth did not have a signific ant impac t on any of our me asur e s of pe r for manc e , safe ty, wor k- family c onflic t or he alth • Pe r for manc e and safe ty me asur e s (inte r pe r sonal inte r ac tions, shooting skills, r isky dr iving be havior s, r e ac tion time , fatigue , and se lf- initiate d de par tme ntal ac tivity) we r e not impac te d by shift le ngth. • T he gr oups did not diffe r with r e gar d to wor k- family c onflic t. 24

Shift le ngth impac t Dur ing the six- month pe r iod in whic h offic e r s we r e assigne d to the e xpe r ime ntal c onditions, did not de te c t diffe r e nc e s ac r oss gr oups: • sic k le a ve ta ke n • stre ss e xpe rie nc e d • inc re a se d c a rdiova sc ula r proble ms • g a strointe stina l proble ms. 25

E ffe c ts of Sc he dule s • 10- hour day is ofte n a pr oduc tivity kille r due to ove r lap • 8- hour day give s maximum assignme nt fle xibility but c an affe c t c ontinuity of supe r vision and te am e ffor t • 12- hour day with 4 platoons and 42- hour we e k ave r age (36 / 48) may be be st alte r native e spe c ially in smalle r age nc ie s. 26

12 Hour Sc he dule – 42 Avg. Hour s Week # SUN MON TUE WED THU FRI SAT 1 Day Team A B B A A B B Nite Team C D D C C D D 2 Day Team B A A B B A A Nite Team D C C D D C C 3 Day Team A B B A A B B Nite Team C D D C C D D 4 Day Team B A A B B A A Nite Team D C C D D C C 27

12- Hour Sc he dule Be ne fits • Sc he dule r e pe ats e ve r y 4 we e ks. • E ac h te am ge ts F r iday, Satur day and Sunday off e ve r y othe r we e k. • T e ams wor k 36 hour s one we e k and the n 48 hour s the following we e k, ave r aging 42 hour s pe r we e k. T his is within F L SA r e quir e me nts and the e xtr a hour s c an be paid at str aight time . • T he se e xtr a 2 hour s pe r we e k c ontr ibute the e quivale nt of one additional offic e r for e ve r y 21 offic e r s. 28

Have you done a De ployme nt vs. Wor kload Analysis? 29

10- Hour Sc he dule : We e kdays PO-Initiated Workloads C/O-Initiated Workloads Total Workloads Deployed Police Officers 16 14 Number of Police Officers Avg. C/O Initiated Manhours of Work: 3.2 12 Avg. Total Manhours of Work: 3.8 10 Avg. Police Officers Deployed: 8.1 8 6 4 2 0 0 2 4 6 8 10 12 14 16 18 20 22 Hours 30

10- Hour Sc he dule : We e ke nds PO-Initiated Workloads C/O-Initiated Workloads Total Workloads 16 Deployed Police Officers 14 Number of Police Officers 12 Avg. C/O Initiated Manhours of Work: 3.4 Avg. Total Manhours of Work: 4.0 10 Avg. Police Officers Deployed: 7.2 8 6 4 2 0 0 2 4 6 8 10 12 14 16 18 20 22 Hours 31

How many polic e do we ne e d? “Offic e r s pe r 1,000 and othe r ur ban myths” by Joe Br ann • Polic y issue drive n by % of non- c ommitte d (fre e ) time • F unc tion of work sc he dule • Wha t the y do with the time is more importa nt tha n how muc h the y ha ve ! 32

How Many Offic e r s? • I n a dditio n to wo rklo a d, re spo nse time s ma y drive sta ffing de c isio ns – pa rtic ula rly in lo w CF S a re a s. • “Pre se nc e ” ma y a lso a ffe c t sta ffing de c isio ns, a g a in in lo w CF S a re a s 33

kload vs. E Wor ve nts 34

Rule of 60 – Polic e Alloc ation • 60% o f the T o ta l Numb e r o f Swo rn Offic e rs sho uld b e a ssig ne d to Pa tro l • No mo re tha n 60% o f a va ila b le ma n- ho urs sho uld b e de dic a te d to wo rklo a d 35