gencode.v16.annotation.gtf is the original ENCODE gtf file and can be obtained here:
ftp://ftp.sanger.ac.uk/pub/gencode/release_16/gencode.v16.annotation.gtf.gz

I used make-tables.R
to extract the transcripts and genes and make two new gtf files.

gencode.v16.annotation.genes.gtf.gz
gencode.v16.annotation.transcripts.gtf.gz

These tables are subsets of the original except for these addition to the attribute column:
1- the first addition is a unique ID (feature_id)
2- the second addition is the row number of the original file corresponding to the entry (original_row)
3- rows were appended to the end to annotate the ENCODE spikeins

Note that the files exon.id.txt, genes.id.txt and transcript.id.txt contain just the feature_ids

You can use the file sanitycheck.R to check if the tables columns and rows match

An example submission is here: exampleSubmission.tgz

Note 1: that there is also an exon table:

gencode.v16.annotation.exon.gtf.gz

in case we want to permit future exon submissions 

Note 2: the files spikein_exon.gtf, spikein_gene.gtf, spikein_transcript.gtf contain just the spike-ins. These have been appended to the end of the annotation files