Our Gemara on Amud Beis (continuing to 31a) discusses a dispute between Bais Shammai and Bais Hillel regarding if an item is declared consecrated for the Bais Hamikdash in error. Bais Shammai holds it remains sanctified, while Bais Hillel holds that since its inception was in error, the sanctification never takes hold.
We have discussed elsewhere in Psychology of the Daf (Eiruvin 13) how thematically Bais Shammai relates to an ideal state, Edenic and Messianic, and therefore often their halakhic positions tend toward a stricter outcome. Shem MiShmuel (Chanukka 35) develops this idea in reference to the famous dispute between Shammai and Hillel over the order of lighting the Chanuka lights. As we know, Hillel says we light from one candle on night one, progressively to eight candles on night eight, while Shammai holds we start at eight and go down to one. The reason Shammai offers is to follow the same style of the counting the Bullocks offered on Succos, which starts at 13 and goes lower by one each day.
Hillel offers a counterargument: We should strive to continuously increase holiness instead of reducing it.
Shammai holds that holiness in the ideal state is initiated by God, and from the moment of this original divine outflow, its influence wanes. This top down approach to spirituality is also represented in the idea that even erroneous sanctifications remain consecrated, because the origination point symbolically assumes that all is holy. Hillel has a more practical view. Humans invoke holiness in this world, thereby causing “celestial arousal” “איתערותא מלעילא”, as maintained by the mystics. Thus, according to Hillel, the enlightenment represented in the lighting of the lamps, increases. Likewise, an erroneous sanctification has no traction. This is why there is a tradition from the Arizal that the halakha will be like Shammai in the world to come (Sefas Emmes, Korach 5647), because even Hillel will agree that since the ideal state has been achieved, the relationship now is starting from intense sanctification.