Rockville City Elections 2003 Election Analysis
Alternatively, you may also view this document by selecting one of the following files.       elec2003.doc (Microsoft Word Document, 253KB)       elec2003.pdf (Adobe Portable Document Format (PDF), 344KB) The vote tallies and calculations used in this report are available in the following file.       elec2003.xls (Microsoft Excel Spreadsheet, 25KB)

Rockville Election Analysis 2003
Roald A. Schrack

Election Day, Tuesday, November 4 was an unseasonably warm sunny day. This was the first election since 1987 to be fully contested, have 2 mayoral candidates and 8 Council candidates for the 4 Council seats. This was the first election since 1954 that did not have candidates running as slates. This was the first Rockville election to use the new touch-screen voting machines.

The political campaign leading up to the election was fairly lackluster with the challenging mayoral candidate, Russ Hamill, throwing his hat in the ring just before the September 5th deadline. There was little political activity during the summer. Activity picked up on September 30 when the first of six public forums took place, with 3 of them rebroadcast on the City cable channel 11. Absentee ballots, which are a reliable indicator of interest in the election started quite slowly, with only 33 received two weeks before the election. As the following table shows, the turnout was a little above average but not as high as the 2001 election. The values given for registration do not include inactive voters. Previous voters refers to those that have voted in at least one of the last four elections. New registration means those that have been added since the last election.

Year	Ballots	Registration	B/R	Previous Voters	New Registration
1987	8056	23,601	.34
1989	6206	23,546	.26	9394	2258
1991	5451	20,648	.26	7916	1741
1993	1449	23,077	.06	9889	4409
1995	4323	22,787	.19	8133	1317
1997	5211	22,660	.22	6968	3299
1999	4690	22,676	.21	6898	1451
2001	6967	24,423	.28	6704	3820
2003	6170	26,198	.24	8372	3004

The average absentee ballots to polling place ballot ratio for the last 5 elections on clear days is .031±.002. Rainy weather reduces the turnout as much as 35%. The predictive quality of the absentee ballot count remained, although many absentee ballots came in the last couple of days.

Year	Absentee Ballots	Absentee Ballots/ Polling Place Ballots
1989	175	.028
1991	168	.032
1993	48	.033
1995	180	.042 rain
1997	156	.031
1999	171	.038 rain
2001	213	.030
2003	192	.031

A spreadsheet shows the results of the election on the last page of this report.

The adjacent graph gives the ratio of ballots to the number of registered voters in a district. This ratio is termed Voter Participation. Data is shown for both this and the last election. Note the wide range in voter participation. The patterns shown persist from election to election reflecting the demographics of the different districts. Similar persistent differences are experienced in the different councilmanic districts in the county. The exceptionally low participation in district 6 is related to three factors. 1) The voters that live in the North Farm part of the district must travel quite a distance to the polling place. 2) There is an ethnic concentration with low voting probability in the garden apartments behind Congressional Shopping Center. 3) The residents in Congressional Towers, as all apartment residents, have lower probability of voting.

District 10 residents are mostly apartment dwellers who participate at a low level. Districts 1,2,8, and 9 where many politically active citizens live historically have had high levels of participation. The dashed lines show the city average voter participation for the indicated years.
This graph shows the absolute ballot difference between the 2001 and 2003 election. Note that only district 10 had more ballots in 2003 than 2001. Even so the voter participation was lower in 2003 than 2001 because of the increase in registered voters there from 1065 in 2001 to 2070 in 2003.
Mayoral Race

The adjacent graph shows the ballot fraction for the two candidates. Giammo lost in district 5 by four votes and was tied by Hamill in 4. For getting in the race so late and with little previous public exposure, Hamill did extremely well. His voters come from two main groups.

1) Personal relationships and the few who actually supported his platform of delaying town center development.
2) Voters that supported Bob Wright in 2001. The next graph shows the votes received by Bob Wright in 2001 and the votes received by Hamill. Note the extreme similarity in the shapes. About 80% of the people who voted for Wright, voted for Hamill.

Giammo got 3926 votes in 2001 and 3736 in 2003, a net loss of 190 votes or 4.8%. The total vote dropped by about 13% so relatively speaking Giammo more than held his own in general level of acceptance.
The Referendum

The 2-year term won easily, receiving about 64% of the vote. The vote by district is shown in the accompanying graph. The vote was probably influenced by some last minute publicity put out by a group entitled the "Rockville Coalition for Responsive Government". It is extremely unusual for so much money to be spent on an advisory referendum as was spent on the two full color expensive mailings. It will be interesting to see who thought it was so important to prevent four-year terms.
The Council Races

The above graph shows the vote fraction received by the eight candidates for council. If you add up all the values for the different candidates plus the Missing Fraction (due to voters not voting for all four council seats) for a district you will get a total of 4. The city uses a different system in which the total is 1. The structureless appearance of the graph is caused by the lack of slates. The relationships between candidates are highly dependent on the district and the strategy used by the supporters of a candidate in that district. Considering the weaker candidates first: Young went from 840 votes in 2001 to 602 in 2003, a percentage drop about twice the overall vote.

Harry Thomas went from 2421 in 2001 to 1619 in 2003, an even greater percentage drop than Young. His only bright spot was in his home district. Nelson, running for the first time, and only living in the city for about a year, got 1486 votes. He will probably do much better if he runs again with more experience and name recognition. It is hard to believe that these candidates motivated any significant voter turnout.

The major races were between the remaining five candidates, four incumbents plus one challenger. Unless the challenger targets one of the incumbents, it is almost impossible for a challenger to win. Consider a voter has four ballots and votes one of them for the challenger. If the voter has no specific incumbent to drop, the three remaining votes will be scattered randomly among the four incumbents. Call this voter A. Consider a second voter, B, that is not voting for the challenger and votes for all the incumbents. If there are 12 type A and 12 type B voters then the final vote tally will be: Challenger 12 votes, Incumbents 21 votes each. If the incumbents are all equally probable of reelection the challenger must get 83% of the vote to win. In this case of 24 total voters, 20 must vote for the challenger. It is rare for the incumbents to be evenly matched, the closest approach to it occurred in districts 6 and 7 where the incumbent votes were all within a 10% range, probably due to intentional slating instigated at the district level. The data does not indicate any attempt at intentional slating of the incumbents in any other districts.

For those candidates that ran in both 2001 and 2003, a comparison was made of their relative vote in the two elections. The only outstanding change was a 55% increase in district 6 for Giammo. This undoubtedly was due to a switch in the proposed vote list handed out at the polling place from Wright in 2001 to Giammo in 2003.

To enhance the probability of winning for a chosen candidate, voters may choose not to vote their allotted four votes. This is called "bullet balloting". The adjacent graph shows the missing ballots for the different districts. There will always be voters who cannot make up their mind on a fourth candidate to vote for, so there are always a few missing ballots. From the number of missing ballots alone it is not possible to tell if .3 means that 10% of the voters only voted for one candidate or 30% of the voters only voted for 3 candidates. If only one candidate is voted for, then the vote totals for that candidate are elevated at the expense of the other candidates. The graph shows that in districts 3 and 8 there was an obvious special effort made at bullet balloting. Going back to the graph showing the vote fractions, it can be seen that Marcuccio was the beneficiary in district 3 and that Robbins was the beneficiary in district 8. In both cases these are the home districts of the candidates and in both cases the vote fractions of all other candidates are depressed. Another indicator of bullet balloting for a specific candidate is the great difference of the winning candidate from the other candidates. In district 3 Marcuccio is 32% higher than the average of the next 3 candidates. In district 8 Robbins is 34% higher than the next three candidates.
Conclusions

1) The absentee ballot count is a good predictor of the final turnout.

2) Voter turnout was not as great as 2001 even though there were more council candidates in addition to a mayoral challenger. Much less money was spent on the 2003 campaign. An analysis of the campaign funding will be made after the December 4^th final financial reports.

3) Giammo seems to have satisfied his supporters and to have retained their votes (and actually increase his relative vote strength).

4) Hamill attracted about 80% of those who voted for Wright in 2001. This was the major component of his strength.

5) Hamill’s platform of cutting back on Town Center development seemed to be ignored by those who endorsed him and voted for him. Was his platform taken seriously, or was he simply seen as a way to show disapproval of Giammo?

6) The lack of slates developed by the candidates does not prevent supporters of candidates from developing local suggested slates in the separate districts.

7) In addition to slates, bullet balloting took place in separate districts for different candidates.

8) Sample ballots handed out to voters at the polling place can strongly affect the vote.

9) New voting machines caused no trouble.

10) The lack of official slates left local activists the opportunity to form slates and/or endorse bullet balloting as they saw fit. Incumbents had a great advantage in this environment.

11) It is extremely difficult for an independent challenger to win against incumbents.

12) The existence of more candidates does not increase turnout unless the candidates excite interest.

 

District	1	2	3	4	5	6	7	8	9	10	Same Day	Absentee	Total
Giammo	352	717	237	290	272	220	465	374	360	292	38	119	3736
Hamill	269	439	159	290	276	101	319	146	180	70	19	64	2332
Mayor write-ins	5	1	0	0	5	1	1	0	0	0	0	0	13
Dorsey	408	713	228	384	332	224	471	265	340	175	24	136	3700
Hall	425	794	241	394	355	230	506	301	369	268	27	132	4042
Hoffmann	386	715	211	291	242	213	460	287	325	226	29	126	3511
Marcuccio	341	624	297	311	272	125	410	220	311	253	32	106	3302
Nelson	178	317	55	107	125	71	180	126	178	105	16	28	1486
Robbins	360	699	187	350	341	210	506	383	269	189	30	100	3624
Thomas	156	294	132	220	287	82	152	76	110	56	9	45	1619
Young	52	91	45	58	85	50	69	57	38	44	7	6	602
Council write-ins	2	6	2	0	12	0	4	1	0	1	0	1	29
Referendum 2	415	751	275	398	397	245	448	322	346	199	39	96	3931
Referendum 4	221	412	120	181	162	77	331	200	202	161	19	87	2173
Registration	2538	4018	1826	2754	2519	2133	3947	2395	1998	2070			26198
Ballots cast	642	1169	404	586	562	325	790	529	549	363	59	192	6170
Mayor sum	621	1156	396	580	548	321	784	520	540	362	57	183	6068
Council sum	2306	4247	1396	2115	2039	1205	2754	1715	1940	1316	174	679	21886
Missed Coun. Votes	262	429	220	229	209	95	406	401	256	136	62	89	2794
Voter Participation	0.25	0.29	0.22	0.21	0.22	0.15	0.20	0.22	0.27	0.18			0.24
Ballot Fraction
Giammo	0.55	0.61	0.59	0.49	0.48	0.68	0.59	0.71	0.66	0.80	0.64	0.62	0.61
Hamill	0.42	0.38	0.39	0.49	0.49	0.31	0.40	0.28	0.33	0.19	0.32	0.33	0.38
Dorsey	0.64	0.61	0.56	0.66	0.59	0.69	0.60	0.50	0.62	0.48	0.41	0.71	0.60
Hall	0.66	0.68	0.60	0.67	0.63	0.71	0.64	0.57	0.67	0.74	0.46	0.69	0.66
Hoffmann	0.60	0.61	0.52	0.50	0.43	0.66	0.58	0.54	0.59	0.62	0.49	0.66	0.57
Marcuccio	0.53	0.53	0.74	0.53	0.48	0.38	0.52	0.42	0.57	0.70	0.54	0.55	0.54
Nelson	0.28	0.27	0.14	0.18	0.22	0.22	0.23	0.24	0.32	0.29	0.27	0.15	0.24
Robbins	0.56	0.60	0.46	0.60	0.61	0.65	0.64	0.72	0.49	0.52	0.51	0.52	0.59
Thomas	0.24	0.25	0.33	0.38	0.51	0.25	0.19	0.14	0.20	0.15	0.15	0.23	0.26
Young	0.08	0.08	0.11	0.10	0.15	0.15	0.09	0.11	0.07	0.12	0.12	0.03	0.10
Total Council fraction	3.59	3.63	3.46	3.61	3.63	3.71	3.49	3.24	3.53	3.63	2.95	3.54	3.55
Missed Council fraction	0.41	0.37	0.54	0.39	0.37	0.29	0.51	0.76	0.47	0.37	1.05	0.46	0.45
Total Council fraction	4.00	4.00	4.00	4.00	4.00	4.00	4.00	4.00	4.00	4.00	4.00	4.00	4.00
Average of Winners	0.615	0.625	0.605	0.615	0.585	0.678	0.615	0.583	0.613	0.645	0.5	0.653	0.605
Standard deviation of W.	0.038	0.032	0.083	0.056	0.045	0.024	0.026	0.083	0.038	0.084	0.029	0.062	0.034