The software that compressed the PDF containing the scanned image mistakenly put the '1' into the background layer instead of the text layer. Then it compressed it together with the rest of the background, using a different algorithm than the foreground. That’s why it looks different. Happens.
Do you have an explanation for the ‘layers’, just curious.
That's all fine and dandy but you are making a supposition. Look at your image... The number one on the background layer is the same size as the text layer that you have shifted, yet the letter M in the time block is clearly a smaller font than the layer you shifted. You have to leave open the possibility that your supposition is wrong and that this document has been manipulated.
As it did in other areas - the 'p' in triplet, the 'R' in Barack, etc. - that make no sense in forging.
It's not a straight scan - the background of the scan has been removed and replaced with the security pattern. There's been too much post processing done to make hard judgements here. If the whole thing had just been a photo, photocopy or straight scan, it would be different and easier, but it's not.