A standard carrier reading wattt meter (think 'Bird') should not move, when the carrier is modulated to 100%. A PEP reading wattmeter (it's my opinion that you can *not* find one that reads accuratly enough) willl only show that the rig -is- being modulated, but not if it is being over-modulated, causing distortion and splatter and being rather un-neighborly to others on near frequencies..
A field strength meter should not move when the carrier is modulated to 100%. It is designed to read AVERAGE carrier voltage. If the final stage has good modulation linearity, the average carrier voltage doesn't change up to 100% modualtion.
A true average-reading wattmeter will read increased output with modulation. The power output includes carrier + sideband power. 1000 watts carrier output should read 1500 (average, or mean) watts out with 100% sinewave modulation.
With a thermocouple rf ammeter in the line, the rf current at the output should show a 22.5% increase with 100% sinusoidal modulation.
Many rf output meters actually read rf voltage output and therefore show no upward kick with modualtion. An rf ammeter is a thermal device which does not average the output, so it will show an upward kick with modulation. Square the current increase to get the percent power increase.
Don k4kyv

