The rankings

I.  Introduction
Ten years or so ago, I created my own livable cities rankings.  My basic goal was to answer the question: what are the best places to live if you don’t want to drive?  I decided to update my page to take account of the radical increase in housing costs in some cities, and to add bikability as an element.
I have chosen to focus narrowly on three factors: crime (which obviously impacts walkability), transit/walkability, and the cost of housing. Crime gets 40 percent, transit/walkability 40 percent, housing 20 percent.
Some factors other people care about that I don’t-  I don’t bother with economic data (because I have never moved anyplace without a job, so the overall economic prosperity of a place doesn’t matter so much to me) or climate (because it is so subjective).
II.    How I rank
A.  Crime
I have one crime test- the murder rate, because homicides are the most verifiable form of crime.  I used to include robbery and burglary, but I have decided to delete them because some cities with high murder rates tend to have pretty ordinary robbery/burglary rates.  It could be that these places are actually not that crime-ridden- but it seems to me that in really dangerous places, people are likely to underreport those crimes.  (The downside of this method, however, is that my measurements are less precise for suburbs where murder is a rare event; if I was trying to create a measurement that works as well for suburbs, I would add other crimes as criteria).  Using homicide only makes a big difference for a few of the worst-off cities: Baltimore, St. Louis, Detroit.  For most cities, including other crimes doesn’t matter very muhc.
To make up for this, I double-weight murder so that it is 30 percent of my overall measurement.  Crime statistics come from this table in the 2015 edition of the FBI’s “Crime in the United States” statistics.
I assign letter grades to places as follows:
Murder: 0=A, up to 2.5 per 100,000 residents = A minus, 2.5-5 = B plus, 5-10 = B, 10-15= B minus, 15-20 = C plus, 20-25 = C, 25-30 = C-, 30-35= D+, 35-40 = D, 40-50 = D-, over 50 = F
B.  Transportation
I use three very different statistics to calculate a place’s friendliness to nondrivers:
  1.  Transit mode share as a percentage of poverty rate- why as a percentage of poverty rate? To exclude places where people take transit only because they are poor.  I rely heavily on the Census Bureau’s American Fact Finder page.
  1. Walkscore (from
  2. Bikescore (also from Walkscore)
I assign letter grades as follows:
Transit market share as percent of poverty rate = under 5% of poverty rate = F, 5-10% = D-, 10-20% = D, 20-33% = D+, 33-50% = C-, 50-75% = C, 75-100% = C+, 100-125% = B-, 125-175% = B, 175-200% = B+, 200%-250%=A-, over 250% -A.
Walkscore- Over 85=A, 80-85= A-, 70-80= B+, 60-70=B, 55-60=B-, 50-55=C+, 40-50=C, 35-40=C-, 30-35=D+, 25-30=D, 20-25=D-, under 20 is F.
Bikescore- same grading system.
Note that I am grading fewer cities than in 2006, because suburbs and small towns rarely have a Bikescore.
C.  Cost of living: I am using two separate tests:
1.  Rents- Regional median rent as a proportion of regional median income, as determined in Geoff Boeing’s recent article.
Why use regional rents instead of municipal rents?  First, because housing markets are effectively regional.  This also means that I will treat cities and their suburbs as identical for housing cost purposes.  Second, I haven’t found any good sources of market rent data that include a large number of central cities; Census data seems suspiciously low, perhaps because it includes government-subsidized housing.
Specifically: if rent is 16-17 percent of income= A
18-19 percent= A-
20-21 percent= B+
22-23 percent= B
24-25 percent= B-
26-27 percent= C+
28-29 percent= C
30-31 percent = C-
32-33 percent= D+
34-35 percent= D
36-37 percent = D-
Over 38 percent = F
2.  Overall housing prices- median housing values (including condominia) as a multiple of median household income.  Data are municipal data from City Data, and tend to reach less surprising (in my opinion) results.  In particular, I grade as follows:
Less than 2-1 (that is, house value less than twice median income) A
Between 2-1 and 2.5-1: A-
2.5 to 3-1: B+
3-1 to 3.5-1: B
3.5-1 to 4-1: B-
4-1 to 5-1: C+
5-1 to 6-1: C
6-1 to 7-1: C-
7-1 to 8-1: D+
8-1 to 9-1: D
9-1 to 10-1: D-
Over 10-1: F
III.  The ratings- I mostly chose major cities (since Boeing’s work is limited to major metros) or suburbs with law schools (since I teach at a law school).  The ratings, in order, are for: murder (remember, double weighted) /transit use/Walkscore/Bikescore/rent/home price.
At the end are two grades.  The first grade averages all these ratings (I assume that A = 95, A- = 92, B+= 88 and so on; an F is a 55).   The second grade ignores cost of living by only weighting the first four (and again, double-weighting murder).
A.        Major cities
Birmingham D       D-        C-             D       B           B+      71.8   65.8
Phoenix         B        D         C        C+            A-          C+      79.8  77.6
 (sorry, not enough data for Tuscon)
Fresno           B         D-        C            C         B-        C+       77.6   76.4
Long Beach  B          C-        B           B         D         D         77.4    82.4
Los Angeles  B           C-        B         B-        D         F          75.6   81.8
Oakland        C          C+       B+         B         D         D-        75.4   80.2
Riverside       B+       D         C           C+       C         C          77.7    78.6
Sacramento  B         D         C             B         B+        C        79.7    79
San Diego     B+       D+       C+         C         C-       D+        76.7    79.4
San Fran-
Cisco              B+       A         A          B+       D         F             82       90.8
San Jose        B+       C-        C+         B-        D         D+       77.2   81.6
Denver           B         C-        B           B+        B        C          82.1     83
Hartford        C-        C-        B+        B         A-        C+       79.8   77.8
Wilmington  D         C-        B+         B         C        B-        76        75
DC                  C          A         B+        B      B             D+           81.5    83.6
Jacksonville  B-         D         D      C-       A-        B             77.5    73.4
Miami            C+         C-        B+      B         F          D-         73.8     80.2
Orlando         B-         D+       C         C+       B-        C+        78        77.2
St. Petersburg B           D         C         B-       B        B           80.3    78.4
Tampa            B            D         C         C+       B         C+          78.6    77.6
Tallahassee   B+        D-        D+       C-       B+       C+         77.7     75.6
 Atlanta          C         C-        C           C+       A-        C+         77.8        75
Honolulu       B+      C+       B        C              C+    F                78           82.8
Chicago         C+         B         B+      B+        C      C+               81.4        83.4
Indianapolis C+          D         D         C         A         B+          77.5         72
Louisville        B-        D         D+       C         A-       B          78.4           74.6
New Orleans D-        D+       B-          B        C+       C         73.2           72.6
Baltimore      F         C+       B            B         B+         B-       75.4            71.6
Boston           B         B         A-          B+       F          D+         79.7            87
Detroit           D-        D+       B-         B-        A-        A         77.8          71.6
Mpls              B-       C         B         A-           B+          C+           83.2       83.4
St. Paul         B         C-        B-      B           B+             B-          82.7       81.8
KC                  C           D         D+       C         A         B+                 77.2       71.6
St. Louis        F           C-        B         B         A        A-                77           70.4
(FUN FACT: KC and St. Louis both plummet to Detroit-like levels without cost of living)
Albuquerque   B        D           C         D+         A-      B-          78.8           77.6
Buffalo           C+       C-        B         A-         C+       A-            82               81
NYC                 B+       A         A       B           F             D-            81.1            90.2
Rochester      C+       D+       B         B-          B         A-             81             78.2
 North Carolina
Charlotte      B         D         D            C-        B+       B               77.8           74.4
Cincinnati      C           D+       C+       C-        A         B                78.2            73.6
(without cost of living, Cincinnati would be at an unattractive 73.6)
Columbus        B            D         C         C         A         B+              81.1              77
(not enough data for Cleveland, as well as smaller cities)
Oklahoma City B-       F         D+        C         A      B+                  78                77.6
Tulsa              B-        D-        C-        C           A         B+              79.5                75.4
 (no crime statistics available for Portland, Or- a city that I would otherwise rank)
Philadelphia  C+     C+       B+       B              C         B-               80.5                81.4
(NOTE: high poverty drags down transit score here)
Pittsburgh    C+       C         B         C             B+       B+                   81                  78.2
(NOTE: mountainous terrain drags down Bikescore).
Rhode Island
Providence    B         D+       B+         B         B-        C+                   81.5              82.2
Nashville      B-           D         D       D+          B       B-                    75.7              72.6
Memphis       C            D-        C-       C          B+      B+                  76.2              71.8
Austin             A-        D+       D+      C+        B        C+                       80.1           79.2
Dallas              B-        D         C         C          B       B                           78.5          76
Ft. Worth      B            F         D+       C          B         A-                      77.8           73.6
Houston         B-         D         C        C           A-       B                       79.5             76
San Antonio    B         D         C-      C          A-        B+                     80.2            76.4
Salt Lake City B+        C-        C+       B        A         C                          83                82
(NOTE: I would have rated Richmond and Norfolk but they don’t have Bike Scores).
Virginia Beach B+     D         D+      C           B+      B-                        79                  76.6
Seattle             B+        B         B+     B            B      C-                            84.2               86.6
(NOTE: Milwaukee is the biggest city without a Bike Score- if you are curious its other rankings are C  D+ B B+ B- if you assume a B Walkscore it gets 80, if lower in the high 70s somewhere)
B.        A few suburbs (mostly with law schools because I was more curious about them) just for comparison’s sake
 Arizona- Tempe (Phoenix suburb)
                                     B+        D+       C+      B+           A-     C+          82.8   81.8
Anaheim   (LA)           B+       D+       C+       C         D        D+         75.7    79.2
Berkeley    (SF)           A-        B         A-        A         D         F           81.2    91.4
Irvine    (LA)                 A-        D         C       B+          D         D          77.5   82.6
Boulder     (sort of a Denver suburb)
                                        A-         C-        B-       A           B       D-        83         86.8
Cambridge (Boston)   B+        A-        A        A         F       D+          82         91.4
IV.  Winners and losers (major cities only): I tried a few different ways of measuring winners and losers.
A. The main formula used above, double-weighting homicide rates and weighting the other five factors equally
Under this formula, the winning formula seemed to be Western, medium-density cities that don’t excel in any way but don’t do horribly in any way either.
The top five:
1.   Seattle 84.2
2.  Minneapolis 83.2
3.  Salt Lake City  83
4.   St. Paul 82.7
5..  Denver 82.2
The bottom five:
1.  Birmingham           71.8
2. New Orleans           73.2
3.  Miami                        73.8
4.  Baltimore and Oakland tied at 75.4
B.  Weighing cost of housing a bit less- what if I eliminated Boeing’s rent measurement, and only used housing values as a measure of cost of  living (thus reducing cost of living from 30 percent of the measurement to 16 percent)?
The top five:
1.  NYC 85.5 (the Boeing measurement disfavors NYC more than the City Data measurement)
2.  San Francisco 84.8
3.  Seattle 84.2
4.  Boston, Buffalo tied at 83.2
The bottom five:
1.  Still Birmingham at 69.6
2.  New Orleans 73
3.   Baltimore 73.3
4.  St. Louis 74
5.  Kansas City 74.3
C.  Ignoring cost of housing:
The top five:
1.  San Francisco   90.8
2.   NYC   90.2
3.  Boston 87
4.  Seattle 86.6 (kind of a surprise, but low poverty and low crime make a big difference)
5.  Washington DC 83.6
The bottom five:
1.  Birmingham 65.8
2. St. Louis 70.4 (kind of a shocker but crime makes a big difference here)
3.  Baltimore, Detroit DC tied at 71.6
D.  Changing transit measurement from “% of people using transit/poverty rate” to “% using transit.” (Grading scale: over 60 percent of commuters walking or using transit to et to work = A, 50-60= A-, 40-50= B+, 30-40=B, 25-30= B-, 20-25=C+, 15-20=C, 10-15=C-, 7-10=D+, 5-7=D, 3-5=D-, under 3 is F.)
Best- Seattle 84, Minneapolis 83.5, Salt Lake City 83, Providence/St Paul/Buffalo 82.7
Worst- Birmingham 71.8, New Orleans 72.6, Miami 74.8, Nashville 75.2, Oakland 75.4
Without cost of living-
Best- NYC 90.2, SF 89.4, Boston 87.6, Seattle 86, Minneapolis 84
Worst- Birmingham 65.8, St Louis 70.4,  Indianapolis 71.4, Kansas City 71.6, Memphis 71.8