[R-sig-ME] Testing whether I need a random effect?

[Christopher David Desjardins]

Hi I have the following model:

m.f <- glmer(NumCitations ~ 1 + Program + ProductType + ProductField + 
(1 | ProductDate), family="poisson",data = data)

And I am wondering how I can test whether or not ProductDate needs to be 
included as a random effect or if I could just run a glm() without 
ProductDate?

Below is the output from m.f. As you can see there variance associated 
with ProductDate is non-zero but should it be included? I presume it's 
wrong to compare an AIC of a glmer() object with a glm() object?

Thanks!
Chris

 > m.f
Generalized linear mixed model fit by the Laplace approximation
Formula: NumCitations ~ 1 + Program + ProductType + ProductField + (1 | 
ProductDate)
Data: lija.ns
AIC BIC logLik deviance
2070 2107 -1026 2052
Random effects:
Groups Name Variance Std.Dev.
ProductDate (Intercept) 0.88876 0.94274
Number of obs: 465, groups: ProductDate, 12

Fixed effects:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 0.91594 0.41865 2.188 0.02868 *
ProgramCETP 0.56466 0.16619 3.398 0.00068 ***
ProgramLSC 1.32624 0.09636 13.763 < 2e-16 ***
ProgramMSP-RETA 0.08312 0.24420 0.340 0.73358
ProductTypePublication -2.61985 0.15060 -17.396 < 2e-16 ***
ProductTypeReport -1.85925 0.10888 -17.077 < 2e-16 ***
ProductFieldSTEM education/research -0.07700 0.38272 -0.201 0.84054
ProductFieldSTEM evaluation 0.85661 0.31292 2.737 0.00619 **
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Correlation of Fixed Effects:
(Intr) PrCETP PrgLSC PMSP-R PrdcTP PrdcTR PFSTEMe/
ProgramCETP -0.074
ProgramLSC -0.109 0.530
PrgMSP-RETA -0.553 0.135 0.201
PrdctTypPbl -0.048 0.390 0.077 0.081
PrdctTypRpr 0.018 0.494 0.041 0.054 0.601
PrdcFSTEMe/ -0.601 -0.205 -0.170 0.574 -0.235 -0.222
PrdctFSTEMe -0.689 -0.200 -0.134 0.626 -0.139 -0.285 0.862